Machine learning (ML)-based mineral prospectivity mapping (MPM): Detecting Iranian plateau high-potential metallogenic zones using geospatial big data

Machine learning (ML)-based mineral prospectivity mapping (MPM): Detecting Iranian plateau high-potential metallogenic zones using geospatial big data | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Machine learning (ML)-based mineral prospectivity mapping (MPM): Detecting Iranian plateau high-potential metallogenic zones using geospatial big data Vahid Teknik, Iman Monsef, Amr Abdelnasser, Abdolreza Ghods This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7730637/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Recent advances in Artificial Intelligence (AI) and Machine Learning (ML) methods have significantly enhanced Mineral Prospectivity Mapping (MPM). These AI-based algorithms offer high capability for regional-scale mapping of underexplored ore deposits. However, there are still significant methodological challenges, particularly in integrating multidimensional, heterogeneous geospatial datasets and handling their inconsistencies with sparse and spatially clustered distributed known mineral ore deposits. The current study presents a novel framework to address these challenges. We developed a training dataset comprising 69 training features derived from geological, geophysical, and lithospheric raster grids. Vector-based geological features were systematically converted into raster grids, where each pixel encodes the minimum distance to the nearest structural and lithological boundaries. Therefore, one can capture the influence of structural and lithological proximity on metallogenic zones. The training target is generated by combining the spatial distribution of seven major metallic ore deposits and converting their spatial locations into a continuous raster grid of spatial density of metallic ore occurrence. Seven ML algorithms with their 24 subtypes are used to predict the spatial density of mineral deposits. Among them, the Ensemble Bagged Trees method showed optimum prediction performance by achieving the lowest Root Mean Square Error (RMSE) and the highest coefficient of determination (R²). The optimized model was applied to calculate a predictive MPM across the Iranian plateau. To pinpoint underexplored high-potential zones, residual spatial density anomalies were calculated by subtracting the observed ore occurrence spatial densities from the predicted prospective grid. The residual spatial density anomalies reveal several promising areas, such as the Malayer-Isfahan Pb-Zn zone along the Zagros suture zone. The residual anomalies show significant potential extended southward of the KaraDagh copper zone in NW Iran. The results also indicate high potential zones in central and eastern Iran, notably near the Bafgh, Nehbandan-Ferdous, and Jiroft-Shahrebabak metallogenic zones. Regional-scale AI-aided regression analysis enhances our understanding of ore deposit distribution across the Iranian plateau. This insight provides a strategic foundation for future national-scale exploration programs by improving efficiency, reducing risk and cost, and narrowing the area of detailed exploration. Artificial Intelligence and Machine Learning Geophysics Geospatial Big Data Machine Learning Mineral Prospectivity Mapping Artificial Intelligence Green exploration Ore Deposits Iran Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Figure 9 Figure 10 Highlights An AI-aided approach was developed for mineral prospectivity mapping. Sixty-nine geological and geophysical variables were used for model training. The predicted spatial density of ore deposits was estimated across the Iranian plateau. Residual spatial density anomalies highlight high-potential underexplored ore deposit zones. The study offers actionable insights for reducing future exploration risks and costs. 1. Introduction The increasing global demand for new metallic ore resources is accompanied by declining discoveries in recent decades, since deposits with surface outcrops have mostly been discovered. It is speculated that unexplored deposits are either concealed beneath sediment cover or obscured by complex tectono-magmatic events (Davies et al., 2021 ). Traditional exploration methods such as geological, geochemical, and geophysical surveying, and drilling are used to identify hidden resources (Zhao, 2007 ; Dentith and Mudge, 2014 ; Adiri et al., 2020 ; Dentith et al., 2020 ). However, the traditional methods are often expensive and time-consuming and face challenges in handling diverse big data formats with typically low resolution and inadequate survey coverage, particularly at the regional scale (Gonzalez-Alvarez et al., 2020 ). To address these challenges and limitations, Mineral Prospectivity Mapping (MPM) has emerged as a fast, cost-effective, and data-driven approach for prioritizing potential metallogenic zones. The MPM approach leverages different geospatial data and insights from known deposits to predict new exploration targets on a regional scale. Formation of ore deposits is governed by various complex geological processes including magmatic differentiation, hydrothermal fluid circulation, sediment accumulation, metamorphic transformation, and supergene enrichment (Bierlein et al., 2006 ; Groves and Bierlein, 2007; Kesler and Simon, 2015 ). Despite substantial progress in understanding these processes, integration of diverse geospatial datasets into a comprehensive mineral exploration model remains challenging (McCuaig et al., 2010 ). The recently introduced artificial intelligence (AI) and machine learning (ML) algorithms have revolutionized geoscientific big data analysis (Bergen et al., 2019 ; Lösing and Ebbing, 2021 ; Woodhead and Landry, 2021 ; Guo and Yang, 2023 ; Li et al., 2023 ; Chukwu et al., 2024 ; Teknik, 2024 ). Recent innovative AI advances are also revolutionizing mineral prospectivity modelling by enhancing processing ability, improving targeting accuracy, and optimizing exploration strategies while minimizing environmental impacts. AI-based methods have been successfully used for uncovering complex relationships between known mineral deposits and associated geological and geophysical datasets (Chen and Wu, 2017 ; Maepa et al., 2021 ; Parsa and Carranza, 2021 ; Parsa and Maghsoudi, 2021 ; N. Yang et al., 2022 ; Yin et al., 2023 ; Zuo and Xu, 2023 ; Li et al., 2024 ; Wake et al., 2024 ). ML-based MPM workflow employs statistical models to predict mineral potential by analyzing key attributes such as lithology, structural patterns, and geophysical anomalies. Then the trained ML models could be used to predict the likelihood of mineral deposits in less explored regions (Holden et al., 2008 ; Carranza and Laborte, 2015 ; Xiong and Zuo, 2018 ; McMillan et al., 2021 ; Benaissi et al., 2022 ). However, the challenges remain because of the scarcity of well-known deposits, their irregular distributions, as well as the complexity of ore formation, which complicate the development of universal models for different ore deposits (e.g., Xiong et al., 2018 ). In this study, we apply AI-based techniques to address the regional-scale exploration challenges on the Iranian plateau. Our approach investigates the statistical relationship between the spatial density of metallic deposits with geological and geophysical data. An integrated metallic mineral dataset was formed by compiling seven ore deposits of iron (Fe), copper (Cu), lead (Pb), zinc (Zn), gold (Au), silver (Ag), and magnesium (Mg). The spatial density of mineral deposits served as a target variable within a series of regression-based learning models, aiming to detect spatial patterns related to metallogenic potential. A novel transformation of complex vector-based geological maps into raster grid formats was introduced in this study to facilitate their integration into a geospatial big data framework suitable for machine learning analysis. We introduced an adaptive AI-aided workflow for regional-scale mineral prospectivity mapping that systematically evaluates the predictive performance of multiple regression algorithms to optimize the detection of the prospective zones across the Iranian plateau. The proposed methodology overcomes the heterogeneity and inconsistency of datasets and provides valuable guidance for prioritizing underexplored metallogenic zones. The approach provides a cost-effective solution with reduced financial risks, thereby facilitating more efficient and targeted mineral resource exploration efforts with minimum environmental impacts. 2. Geological framework of the Iranian plateau The Iranian plateau represents a complex and dynamic geological setting, where shaped by a series of temporally and spatially diverse geodynamic and tectonic events. Its evolution is closely linked to the sequential opening and closure of the Paleo- and Neo-Tethyan oceans, events that played a crucial role in the region's tectonostratigraphic architecture and metallogenic development (Stampfli and Borel, 2002 ). Therefore, Iranian plateau is composed of continental blocks that are separated by Tethyan oceanic suture zones. The convergence between Arabian and Eurasian plates, accompanied by the progressive accretion of multiple continental microplates along the southern margin of Eurasia, where it has fundamentally shaped the present-day configuration of the Iranian plateau (Stocklin, 1968 ; Stöcklin, 1974 ; Berberian and King, 1981 ; Alavi, 1996 ). The ore deposits' formation, their types and spatial distribution are closely associated with the magmatic history across the Iranian plateau, especially where the various phases of crustal extensions and compressions within the Tethyan orogeny, extending from the Early Palaeozoic to the Cenozoic (Stampfli, 2000 ; Richards, 2015 ). Many of the major crustal-scale lineaments were reactivated during Mesozoic and Cenozoic tectonic events and facilitated both extensive crustal extension and exhumation intrusions (Shahabpour, 1999 ; Moritz et al., 2006 ; Meshkani et al., 2013 ; Bagheri, 2015 ). The exposures of Neoproterozoic-Early Cambrian crystalline basement rocks in the Iranian plateau and the trend of magmatic belts provide key controls on the distribution and localization of mineral deposits (Meshkani et al., 2013 ). Extensional tectonics associated with the rifting of the Neo-Tethys ocean led to the detachment of the Cimmerian terranes from northern Gondwana during the Permian to Triassic. These Cimmerian terranes include the Anatolide-Tauride, Central Iran, Tibet, and Indochina. The northward drifting of these terranes ended with the event of their collision with the southern margin of Eurasia from the Late Triassic to the Early Jurassic, which led to the formation of the Paleo-Tethyan suture zone (Stampfli, 2000 ; Stampfli and Borel, 2002 ; Richards, 2015 ). During the Mesozoic and Cenozoic, the subduction of the Neo-Tethys ocean beneath the Central Iran blocks formed fore-arc basins, extensive magmatism, and back-arc basins throughout the Iranian plateau (e.g., Monsef et al., 2022 ). One of the most significant magmatic arcs within the Iranian plateau is the Urumieh-Dokhtar magmatic arc (UDMA), which is marked by a subduction-related magmatism caused by the subduction of the Neo-Tethys ocean beneath the Central Iran block. The Urumieh-Dokhtar magmatic arc mostly formed by pronounced magmatic flare-ups throughout the Eocene to the Oligocene (ca. 55 − 25 Ma) (Verdel et al., 2011 ; Chiu et al., 2013 ). The magmatic flare-ups of UMDA were followed by widespread post-collisional extensional magmatism during the Neogene, which is attributed to the slab break-off or lithospheric delamination (e.g., Allen et al., 2013 ). The composite tectonic unit of the Central Iranian block consists of three major blocks of Lut, Tabas, and Yazd (Fig. 1 ). Eastern Iran includes the Sistan suture zone and Lut block. Eastern Iran is characterized by intense Tertiary strike-slip faulting, widespread magmatism, and N-S trending ophiolitic belts (e.g., Omidianfar et al., 2020 ). Subduction-related magmatism associated with these tectonic processes has contributed significantly to the metallogenic evolution of Eastern Iran (Arjmandzadeh et al., 2011 ; Alaminia et al., 2013 ; Richards, 2015 ). In the northeast of Central Iran and north of Eastern Iran, the E-W trending Sabzevar magmatic-ophiolitic belt extends about 700 km along the northern side of the Dorouneh fault. This belt is bound to the Kopeh-Dagh zone (Eurasian plate) in its north. The Kopeh-Dagh Mountains mark the most northeastern edge of the deformation zone of the Arabia-Eurasia collisional zone in the Iranian plateau. The Kopeh-Dagh is marked with a ~ 10 km deep folded Mesozoic-Tertiary sediment sequence. The Sistan and Sabzevar ophiolites represent the evidence of Neo-Tethys back-arc basins and serve as an important lithological marker delineating the boundaries of major tectonic units (Stöcklin, 1968 ; Richards, 2015 ). During the Neogene, the Arabian plate collided with the Central Iran block, leading to the formation of the Bitlis-Zagros suture zone (Stöcklin, 1968 ; Berberian and King, 1981 ; Boulin, 1991 ; Stampfli and Borel, 2002 ). The continuous northward convergence of the Arabian plate gave rise to the Zagros fold and thrust belt with an extensive NW-SE-trending orogenic belt in the south of the Zagros suture zone. The Zagros orogenic system remains tectonically active, particularly where the Main Recent Fault (MRF) and the Main Zagros Thrust (MZT) consume part of the regions' ongoing continental convergence (Stampfli and Borel, 2004 ; Richards, 2015 ; Berberian, 1995; Sepehr and Cosgrove, 2004) (Fig. 1 ). The Zagros suture zone lies approximately along MRF and MZT. The Zagros ophiolites are discontinuously exposed along the Zagros suture zone. The ophiolites can be divided into two parallel belts of the inner and outer Zagros ophiolitic belts. In the southwest periphery of the Central Iran block, the inner Zagros ophiolitic belt comprises Nain, Dehshir, and Baft ophiolites. The outer Zagros ophiolitic belt, along the Main Zagros Thrust, comprises Kurdistan, Kermanshah, Neyriz, and Hajiabad ophiolites (e.g., Monsef et al., 2018 ). The southeastern continuation of the Zagros fold and thrust belt identifies the Makran accretionary prism that extends for ~ 900 km from southeastern Iran to southeastern Pakistan. In this region, the last piece of Arabian oceanic lithosphere is currently being subducted beneath the Makran accretionary complex in the south of Iran and Pakistan. The Makran subduction zone, which exhibits an anomalously low level of magmatism, is associated with the development of the Cenozoic Jazmurian Basin, interpreted as an old back-arc basin (Glennie et al., 1990 ; Shahabpour, 2010; Penney et al., 2017 ; Burg, 2018 ; Teknik et al., 2024 ). The Jazmurian basin is covered with a thick sedimentary cover reaching its maximum thickness of ~ 20 km in its eastern edge (Enayat and Ghods, 2023 ; Mehrdar et al., 2025 ). The Zagros ophiolites continue southeast, toward the Makran ophiolites. The Makran ophiolites can also be divided into two distinct belts, consisting of the inner and outer Makran ophiolitic belts (e.g., Monsef et al., 2019 ). 3. Ore deposits of the Iranian plateau Different world-class ore deposits occurred in the Iranian plateau, making the Iranian plateau one of the most important metallogenic zones in western Asia and central Tethys (Fig. 2 and Table 1 ). The most prominent ore deposits on the Iranian plateau are iron (Fe), copper (Cu), lead-zinc (Pb-Zn), and gold (Au) deposits. This is indebted to the complex geologic setting of the plateau, manifested by a series of tectonic, magmatic, hydrothermal, and metamorphic events along the major active geological boundaries. These processes have facilitated the formation of distinct metallogenic zones. Each zone is particularly associated with the distinct tectonic structures such as faults, shear zones, and suture boundaries, as well as geodynamic events such as subduction, continental collision, and post-collisional extension. The investigation of ore deposits was done using various methods such as remote sensing, geochemical analyses, and structural surveys (e.g., Zarasvandi et al., 2005 ; Golmohammadi et al., 2015 ; TaleFazel et al., 2019 ). The influence of the geodynamic processes on the genesis of the Iranian ore deposits is addressed by comprehensive geological studies. In this context, the evolution of Tethyan oceans and associated tectono-magmatic activities caused the distribution of ore deposits within the Iranian plateau. Indeed, the sequential divergent and convergent events of the Tethyan oceans provide a suitable geological framework for the formation of a variety of ore deposits (e.g., porphyry, skarn, epithermal, MVT, VMS, and SEDEX). Igneous rocks, including volcanic, sub-volcanic, and plutonic bodies, crystallized from the magmas that derived from the partial melting of the upper mantle and then differentiated in crustal magma chambers. The metalliferous hydrothermal fluids exsolved from the evolved melts caused Fe, Cu, Pb-Zn, and Au mineralisation in the form of massive ores, breccias, veins, and veinlets at the shallow levels. The subduction of Tethyan oceans beneath the Iranian continental fragments produced arc-related and extensional back-arc basin magmatism with mostly calc-alkaline to alkaline affinities. After the collision between the Arabia and Central Iran or Lut and Afghan blocks, the lithospheric delamination and subsequent asthenospheric upwelling led to decompression melting of previously metasomatized sub-continental lithospheric mantle and the generation of high-K alkaline to shoshonitic volcanism and plutonic bodies. Porphyry deposits in Iran are intrusion-related deposits that are products of magmatic-hydrothermal activity at shallow crustal levels in subduction and post-collisional tectonic settings. Primarily copper, but also molybdenum and gold occur closely related to epizonal intrusions of porphyric magmatic rocks (e.g., Zarasvandi et al., 2005 , 2019 ; Hezarkhani, 2006 ; Aghazadeh et al., 2015 ). Skarn-type deposits in Iran are formed by metasomatic replacement of carbonate rocks by hydrothermal fluids derived from granitoid bodies in subduction and collisional tectonic settings. They are characterized by the presence of copper and iron (e.g., Golmohammadi et al., 2015 ; Hassanpour and Rajabpour, 2020 ). Epithermal deposits in Iran, ranging from high-sulfidation to low-sulfidation types, are connected with terrestrial volcanism that formed by near-surface magmatic-hydrothermal processes circulating through fault systems and veins. They are often hosted within volcanic and volcaniclastic rocks remarkable as a major source of gold, silver, copper, lead, and zinc (e.g., Mehrabi et al., 2016 ; TaleFazel et al., 2019 ). Mississippi Valley-type (MVT) deposits in Iran are a type of epigenetic carbonate-hosted sulfide ore deposit, primarily known for their lead and zinc mineralization. These deposits are characterized by their migration of ore-forming fluids within fault and karst systems in a collisional tectonic setting (e.g., Rajabi et al., 2013 ; Qaderi et al., 2024 ). Volcanogenic massive sulphide (VMS) deposits in Iran are formed by submarine volcanic activity, or more precisely, seafloor hydrothermal activity related to the arc/intra-arc rifts and back-arc basins. The most important metals in VMS are copper, lead, and zinc, with trace contents of gold and silver that are hosted in volcano-sedimentary succession (e.g., Hajsadeghi et al., 2018 ; Mousivand et al., 2018 ). Sedimentary exhalative (SEDEX) deposits in Iran are very similar in genesis to the VMS deposits that are primarily known for containing often lead and zinc with subordinate silver and iron formed by the precipitation of metal sulfides from hydrothermal fluids onto the seafloor sediments of arc/intra-arc rifts and back-arc basins (e.g., Maghfouri and Hosseinzadeh, 2018 ; Maghfouri et al., 2025 ). Accordingly, these deposits within the Iranian plateau are formed in an integral ore-forming system and have a genetic link with magmatic-hydrothermal processes. These magmatic-hydrothermal activities have a relationship to the evolution of the Tethyan oceans. Central Iran, particularly the Bafgh region, is recognized as one of the most metallogenically significant zones in the Tethys belt, especially for its exceptional concentration of iron ore deposits. The region hosts some of the largest iron ore bodies within the Iranian plateau, such as the Choghart and Sechahoon deposits. These gigantic deposits are primarily associated with Neoproterozoic metamorphic and igneous complexes (Ghorbani, 2013a ; Korehie et al., 2019 ). High aeromagnetic anomalies around the Bafgh region imply a high potential for further iron metallogenic zones (e.g., Torab and Lehmann, 2007 ). Iron ore spatial distribution extends beyond Central Iran into the Sanandaj-Sirjan zone, Eastern Iran, and the Urumieh-Dokhtar magmatic arc. It includes a variety of deposit types such as magmatic, skarn, volcanogenic, and sedimentary types (e.g., Stosch et al., 2011 ; Nabatian et al., 2015 ; Hassanlouei and Rajabzadeh, 2019 ). Numerous porphyry and skarn-type cupper ore deposits are predominantly associated with the Urumieh-Dokhtar magmatic arc. Significant clusters of porphyry copper deposits are found in the Jiroft-Shahrebabak zone (Fig. 2 d), where the world-known porphyry copper deposits of Sarcheshmeh and Meiduk have been discovered (Fig. 2 a). A similar tectonomagmatic activity and emplacement of magmatic intrusions along Urumieh-Dokhtar magmatic arc formed the Mazraeh and Sungon copper deposits (Fig. 2 a), together with a cluster of minor deposits along the KaraDagh zone in northwestern Iran (Zarasvandi et al., 2007 ; Hezarkhani, 2008 ; Mollai et al., 2009 ; Asadi et al., 2014 ; Kheyrollahi et al., 2018 ). Lead-zinc deposits are widespread across the Sanandaj-Sirjan zone and more specifically the Malayer-Isfahan zone. Other major lead-zinc metallogenic zones are situated in Central Iran and the Alborz regions (Ghorbani, 2013b ). Considering the diverse tectono-magmatic and metamorphic events, the various metallic ore deposits have been categorized based on their spatial distributions and types (Ghorbani, 2013b ). The gold deposits on the Iranian plateau are relatively less explored. However, the distribution of gold deposits is primarily extended within the Sanandaj-Sirjan metamorphic zone as well as along the Urumieh-Dokhtar magmatic arc, Azerbaijan-Alborz, and Eastern Iranian magmatic belts. These regions host epithermal, porphyry-related, and orogenic deposit types, reflecting the complex geodynamic evolution of the Tethyan metallogenic belt across the Iranian plateau (e.g., Moritz et al., 2006 ; Richards et al., 2006 ; Daliran, 2008 ; Geranian et al., 2016 ). The silver deposits, similar to gold deposits, are typically connected to the copper metallogenic zones. The spatial distribution of the magnesium mostly follows the trends of the Sistan suture zones (Fig. 2 a). In this study, the metallogenic zones delineated by Ghorbani ( 2013a ) have been refined and updated by calculating the spatial density of mineral deposits (Fig. 2 d). Regions exhibiting high spatial density (> 0.2 num/sq km) of mineral deposits have been identified. Consequently, the major metallogenic zones on the Iranian plateau are: (1) KaraDagh; (2) Takab-Tarom-Hashtjin; (3) Malayer-Isfahan; (4) Kashan-Natanz; (5) Toroud; (6) Anarak; (7) QaleBala; (8) Bafgh; (9) Khaf; (10) Nehbandan-Ferdous; (11) Jiroft-Shahrebabak; and (12) Kahnuj-Fanuj. Table 1 List of major known Cu, Fe, Au, and Pb-Zn ore deposits throughout the Iranian plateau (after Meshkani et al., 2013 and references therein). The selected deposits here are grouped for qualitative evaluation of the AI-based models. Abbreviations: MVT, Mississippi Valley type deposit; VMS, Volcanogenic massive sulphide deposit; SEDEX, Sedimentary exhalative deposit. Deposits Long Lat Host/country rocks Stratigraphic age Major Commodity Other accessory minerals Genetic type Tonnage and grade Masjed daghi 46.936 38.875 Granite and andesitic rocks Oligo-Miocene Cu Au Porphyry 10 Mt − 0.7% Cu, 0.5 g/t Au Mazraeh 46.983 38.625 Volcano-sedimentary rocks, granite, and limestone Oligo-Miocene Cu Pb, Zn, Au, Ag Skarn 1 Mt − 1.7% Cu, 0.3 g/t Au Sungon 46.717 38.700 Monzodiorite and volcano-sedimentary rocks Oligo-Miocene Cu Mo, Pb, Zn, Au, Ag Porphyritic-skarn 290 Mt − 0.76% Cu, 0.015% Mo Enjerd 46.933 38.683 Andesite, granite, and limestone Cretaceous, Oligo-Miocene Cu Au, Fe Skarn 300Mt − 0.85% Cu Astamal 46.375 38.567 Andesite and tuff Oligo-Miocene Cu Mo, Pb, Zn Vein > 1 Mt Jarou 50.550 35.708 Andesite and andesitic basalt Eocene-Oligocene Cu Pb, Zn Vein > 1 Mt Veshnaveh 50.983 34.233 Andesitic basalt and tuff Eocene Cu Vein > 1 Mt Deh madan 51.083 31.600 Limestone, dolomite, and sandstone Cambrian Cu Zn, Pb, Co MVT > 1 Mt Meskani 53.450 33.325 Trachyandesite and basalt Eocene Cu Ni, Co, U, Bi, Au, Pb, Zn, Ag Hydrothermal (volcanogenic) ~ 1 Mt.- 2%Cu, 002% Ni, 15 g/t Ag Talmesi 53.450 33.383 Trachyandesite and basalt Eocene Cu Ni, Co, U, Bi, Au, Pb, Zn, Ag Hydrothermal (volcanogenic) ~ 1 Mt − 2.2% Cu, 002% Ni Ali abad 48.983 36.508 Porphyritic granodiorite and tuff Oligo-Miocene Cu Mo Porphyry 40 Mt − 0.7% Cu, 0.005% Mo Darreh zereshk 53.842 31.567 Granodiorite, andesitic tuff and limestone Eocene-Miocene Cu Porphyry 23 Mt − 0.68% Cu, 0.01% Mo Chah mousa 54.867 35.475 Andesitic tuff and lava Paleogene Cu Vein > 1 Mt Taknar 57.783 35.367 Rhyolite and schist Late Paleozoic Cu Ag, Au, Pb, Zn SEDEX > 2 Mt − 3% Cu, 1.5% Zn, 1.%Pb Qaleh zari 58.955 32.362 Andesitic to basaltic lava and tuff Eocene Cu Fe, Au, Ag, Pb, Zn Hydrothermal 1.3 Mt − 3% Cu, 2 g/t Au, 30 g/t Ag Meiduk 55.200 30.467 Andesitic basalt and pyroclastic rocks Miocene Cu Porphyry 500 Mt − 0.83% Cu, 0.01% Mo Darreh-zerreshk 55.917 29.883 Volcano-sedimentary rocks and porphyritic diorite Miocene Cu Mo, Pb, Zn Porphyry > 1 Mt 0.64%, 0.004% Mo Sarcheshmeh 55.867 29.950 Porphyritic granodiorite and andesite Eocene-Miocene Cu Mo, Au Porphyry 1200 Mt − 0.7% Cu, 0.03% Mo, 0.08 g/t Au, 3 g/t Ag Chahar gonbad 56.183 29.592 Porphyritic quartz diorite and andesitic tuff Eocene-Oligo-Miocene Cu Pb, Zn, Au, Ag Hydrothermal 3 Mt − 1.67% Cu Sheikh aali 56.758 28.133 Pillow lava and volcanic rocks Upper Cretaceous Cu Au VMS > 1 Mt − 2% Cu, 64 g/t Au Rameshk 58.817 26.817 Gabbro, andesitic basalt and limestone Upper Cretaceous-Lower Paleocene Cu VMS? > 1 Mt Anguran 47.406 36.628 Limestone, micaschist, and marble Proterozoic Pb, Zn Ag Carbonate-hosted 22 Mt − 24% Zn, 6% Pb Alam kandi 47.283 36.717 Schist, marble, quartzite, and tuff Proterozoic Pb, Zn Cu Carbonate-hosted ~ 1 Mt − 7% Zn, 3% Pb Ahangaran 48.991 34.186 Sandy dolomite, quartzite and shale Lower Cretaceous Pb, Zn Fe, Ba, Cd, Ag MVT 2Mt − 3% Pb, 1% Zn, 200 g/t Ag Emarat 49.603 33.856 Limestone and dolomite Cretaceous Pb, Zn Cu, Ag, Cd MVT 10 Mt − 2% Pb, 3% Zn Dona 51.450 36.165 Limestone and dolomite Permian Pb, Zn Ba, Ag Carbonate-hosted 6.5 Mt − 5% Pb, 1% Zn, 150 g/t Ag Darreh noqreh 50.217 33.525 Limestone, pyroclastic and volcanic rocks Cretaceous Pb, Zn Cu, Ag MVT ~ 1 Mt − 21% Pb, 2% Zn, 150 g/t Ag Lakan 50.648 33.109 Silicified limestone and shale Cretaceous Pb, Zn Cu, Ag, Ba SEDEX 5 Mt − 3% Zn, 4.5% Pb Hosein abad 50.983 32.958 Black shale and sandstone Jurassic Pb, Zn SEDEX 2 Mt − 1% Zn, 4% Pb Anjeereh tiran 51.125 32.745 Dolomite, limestone, and shale Cretaceous Pb, Zn Cu, Ag, Cd, Sb MVT 1.5 Mt − 4% Zn, 1% Pb Irankuh 51.625 32.500 Dolomite, limestone, and shale Cretaceous Pb, Zn Ag, Cd MVT 17 Mt − 11% Zn, 2.5% Pb Kuh-e-surmeh 52.517 28.500 Dolomite, limestone, and sandstone Lower Paleozoic Pb, Zn MVT ~ 1 Mt − 17% Zn, 2%Pb Nakhlak 53.839 33.564 Limestone, shale, sandstone, and marl Middle Triassic-Upper Cretaceous Pb, Zn Ag MVT 3 Mt − 5% Zn, 75 g/t Ag Moujen 54.658 36.525 Limestone Permo-Triassic Pb, Zn Fe MVT ~ 1 Mt − 2% Pb, 4% Zn Khan jar 54.558 35.367 Dolomite and limestone Cretaceous Pb, Zn Ag Carbonate-hosted 1 Mt − 20% Pb, 4% Zn, 300 g/t Ag Chah sorb 56.617 34.050 Dolomite and limestone Middle Triassic Pb, Zn Ag Carbonate-hosted ~ 1 Mt − 5% Zn, 2.5%Pb Ozbak kuh 57.117 34.667 Limestone, shale, and sandstone Paleozoic Pb, Zn Carbonate-hosted 2 Mt − 10% Zn, 4%Pb Mehdi abad 55.025 31.483 Limestone, dolomite, and schist Cretaceous Pb, Zn Fe, Ba MVT 218 Mt − 7% Zn, 2.3% Pb, 51 g/t Ag Koushk 55.775 31.733 Black shale Proterozoic Pb, Zn Ag SEDEX 5 Mt − 15% Zn, 3%Pb Chah mir 56.042 31.65 Black siltstone Late Cambrian Pb, Zn SEDEX ~ 1 Mt − 6% Zn, 3.5%Pb Aghdarreh 47.017 36.667 Silicified limestone Lower Miocene Au Sb, As, Hg Sediment hosted > 5 Mt − 4.5 g/t Au Zarshuran 47.133 36.725 Dolomitic limestone and black shale Upper Proterozoic-Lower Cambrian Au As, Sb, Hg, Zn Sediment hosted 12 Mt − 7.9 g/t Au Touzlar 47.466 36.833 Volcanic rocks Oligo-Miocene Au Low sulfidation epithermal ~ 1 Mt − 3.1 g/t Au Kervian 46.100 36.133 Meta-volcanic and schist Mesozoic Au Orogenic gold ~ 1 Mt − 3 g/t Au Alut 45.597 36.147 Quartz schist Mesozoic Au Orogenic gold ~ 1 Mt − 2.5 g/t Au Sari gunay 48.092 35.183 Porphyritic micro-diorite and rhyolite Tertiary Au Sb, Cu, As, Hg High sulfidation epithermal 108 Mt − 2.3 g/t Au Astaneh 49.325 33.867 Micro-granite Triassic-Jurassic Au Cu, W Intrusion-related gold ~ 1 g/t Au Muteh 50.608 33.667 Schist and meta-rhyolite Upper Proterozoic-Lower Cambrian Au Cu Epithermal 8 Mt − 3.2 g/t Au Zarrin 54.625 32.675 Alluvium Quaternary Au W Placer 1 Mt - < 1 g/t Au Kuh-e-zar 54.650 35.467 Alluvium Quaternary Au Placer ~ 1 Mt − .05 g/t Au Gandi 54.633 35.317 Volcano-sedimentary rocks Eocene Au Cu, Pb, Zn, Ag Intermediate epithermal ~ 1 Mt − 5 g/t Au Zar mehr 58.927 35.174 Andesite and granodiorite Eocene Au IOGC 1.5 Mt − 4 g/t Au Zartorosht 57.211 28.219 Greenschist Paleozoic Au Orogenic gold type > 1 Mt − 3 g/t Au Shahrak 47.828 36.420 Limestone, rhyodacite, and andesite Oligo-Miocene Fe Volcanogenic 100 Mt − 57% Fe Shams abad 49.725 33.817 Sandy dolomite and shale Cretaceous Fe Mn, Cu Volcanogenic 30 Mt − 47% Fe, 4% Mn Sangan 60.400 34.408 Granodiorite, limestone and schist Proterozoic Fe Skarn 900 Mt − 47% Fe Robat posht badam 55.567 32.958 Gneiss, amphibolite, and marble Triassic Fe Magmatic > 1 Mt Chador Malu 55.500 32.300 Meta-syenite, schist, and rhyolite Proterozoic Fe Magmatic 450 Mt − 56% Fe Sechahoon 55.643 31.718 Metasomatic granite, andesite, and tuff Proterozoic Fe Magmatic 132 Mt − 35% Fe Choghart 55.467 31.700 Syenite, schist, and rhyolite Proterozoic Fe Mn Magmatic 350 Mt − 55% Fe Gol Gohar 55.083 29.267 Schist, marble, and quartzite Proterozoic Fe Skarn 1200 Mt − 55% Fe Tang-E-Zagh 56.017 27.950 Limestone and marl Proterozoic Fe Mn Sedimentary 5 Mt − 42% Fe, 1% Mn 4. Training dataset The construction of the training database is conducted through the following multi-stage workflow: 1) Compilation of publicly available geological and geophysical datasets, which are collected from authoritative references (Table 2 ); 2) Feature enhancement through the use of edge-detection and gradient-based filters, which provide additional training features. These features highlight geological contrast and structural delineation; 3) Target definition is a parameter that represents a metallogenic proxy. Here, the training target was generated based on the spatial density of mineral occurrence. This parameter illustrates the intensity and distribution of metallogenic intensity (Fig. 2 ); 4) The training dataset was formed by compiling all training input databases (Table 2 ), which were sampled ore deposit locations (collected data in step 1). The training dataset represents the relevance of the geological or geophysical features to the training target. The original compiled training dataset consists of 2351 sample points. Before any process, a certain data quality checking and data cleaning are required for preparing raw data for model training. To have high-quality data to make reliable predictions, the following steps are needed. (i) identifying and handling missing data, (ii) removing duplicates, (iii) handling outliers with unrealistic values of training features. If for one of the data points, all feature training data sets does not exist, that data point is removed. After visually checking the raw dataset and removing the problematic data points, the cleaned training data set with 2037 observation points was used to train the model. The suggested database workflow supports a geologically informed and data-consistent training framework, crucial for generating robust and interpretable mineral prospectivity models throughout the Iranian plateau. Implementing ML-based mineral prospectivity mapping needs sophisticated methods for extracting meaningful and geologically plausible patterns from training datasets. A crucial prerequisite for achieving accurate model performance is ensuring compatibility and consistency in the format of training and the target features. The training features comprise diverse geoscientific datasets, each providing complementary insights into mineral systems. The widely used training features are geological data (e.g., Carranza, 2009 ; Li et al., 2021 ), geophysical data (e.g., Chen et al., 2015 ; Anderson et al., 2017 ), geochemical data (e.g., Soloviev et al., 2019 ; Xiong and Zuo, 2020 ), and remote sensing data (Bruzzone et al., 2006 ; Beygi et al., 2021 ; Shirmard et al., 2022 ). In the current study, the training features are geophysical measurements (Fig. 4 ; Table 2 ), mapped geological units (Fig. 5 ; Table 2 ), and spatial variation of the major lithospheric structures (Fig. 6 ; Table 2 ). The selected training feature datasets (Table 2 ) are underpinned by well-established geological principles and empirical evidence from the Iranian plateau’s mineral systems. Each variable serves as a proxy for ore-controlling geological processes. For instance, proximity to igneous rock units indicates potential for porphyry Cu–Mo–Au and skarn-type Fe mineralization, especially within the Urumieh-Dokhtar magmatic arc, where Tertiary intrusions have been genetically linked to major deposits such as Sarcheshmeh and Meiduk (Aghazadeh et al., 2015 ). Similarly, specific lithological units such as rhyolite and dacite are associated with high-sulfidation epithermal gold systems and VMS deposits due to their arc volcanic origins. Mafic-ultramafic rocks (e.g., basalt, gabbro) present in ophiolitic belts act as host rocks for Cyprus-type VMS Cu–Zn deposits, like those found in SE Iran (Hajsadeghi et al., 2018 ). Structural features such as faults and fault density layers are included due to their role as conduits for hydrothermal fluids and their spatial association with Au and Cu deposits, particularly in the Sanandaj–Sirjan Zone (Aliyari et al., 2012 ). Moreover, stratigraphic layers reflecting Paleozoic and Mesozoic age units capture the metallogenic potential of formations that host Sedex Pb–Zn and MVT deposits (e.g., Koushk, Chahmir). These variables collectively reflect the regional metallotectonic framework shaped by Neo-Tethyan subduction and subsequent orogenesis, offering a geologically sound basis for mineral potential modelling. The geophysical data, including aeromagnetic and gravity datasets, are generally provided as continuous raster grids. These raster grids are composed of equally spaced square grids (pixels) and require minimal pre-processing for format alignment. For this study, all geophysical features were harmonized to a raster grid with a 2 km cell size, representing the resolution of spatial analysis. It is expected that the gravity and magnetic field and their derivatives highlight the effects of crustal scale faults and lineaments on the spatial distribution of ore deposits. The geological features are provided primarily in vector formats of point data (e.g., ore deposits), polylines (e.g., faults and tectonic boundaries), and polygons (e.g., lithological units). Converting these to raster format enhances the visualization of their spatial relationships. This raster grid allocates each node or pixel the minimum distance to the nearest geological unit, facilitating a uniform and spatially obvious analysis. The transformation was implemented by calculating the Euclidean distance from each raster cell to the closest geological feature. The Euclidean distance is a standard metric for distance estimation of features in the GIS platforms, especially for MPM approaches. Thereby, the value of each point in the transformed raster grids represents the spatial proximity as a continuous variable throughout the grid. The distribution of known mineral deposits across the study area is spatially irregular and sparse, often with a lack of comprehensive negative (non-mineralized) training examples. This heterogeneity and discontinuous dataset pose challenges for supervised ML approaches. To address this issue, a spatial density grid estimation of mineral deposits was applied to transform the discrete mineral deposit occurrence data into a continuous surface representing metallogenic intensity. The resulting spatial density grid has a raster grid format with a 2 km cell size to align with the training target. Table 2 List of the geological and geophysical training feature layers. Vector data (MD* raster grids), (This study) Raster grids Igneous rock units (IG_RU) Magnetic anomaly & derivatives : Ophiolite Total magnetic intensity (TMI; Teknik et al, 2017) Fault Variable reduced-to-pole (VRTP) TMI, (This study) Rhyolite 1st Vertical gradient of TMI VRTP, (This study) Gabbro 1st x_horizontal gradient of TMI VRTP, (This study) Granite 1st y_horizontal gradient of TMI VRTP, (This study) Diabase Analytical signal of TMI VRTP, (This study) Diorite ACMS** Basalt Gravity anomalies (Pavlis et al., 2012): Dacite Free Air gravity anomaly (FAA) Cretaceous Analytical signal of FAA, (This study) Devonian Bouguer gravity anomaly (BG) Early Cretaceous Analytical signal of BG, (This study) Early Jurassic Gravity gradients components *** Early-Middle Triassic Gxx Early Paleozoic Gxy Eocene Gxz Eocene-Oligocene Gyy Late Cretaceous Gyz Cretaceous-Early Paleocene Gzz Late Eocene-Oligocene Elevation model (Amante and Eakins, 2009 ) Late Jurassic Digital elevation model (DEM) Late Jurassic-Early Cretaceous 1st Vertical gradient of DEM, (This study) Late Paleozoic 1st x_horizontal gradient of DEM, (This study) Middle Jurassic 1st y_horizontal gradient of DEM, (This study) Miocene Analytical signal of DEM, (This study) Miocene-Pliocene Lithospheric major structures ( Irandoust et al., 2022 ): Oligocene Average Vs**** of upper crust, (This study) Oligocene-Miocene Average Vs**** of lower crust, (This study) Ordovician-Silurian Average Vs**** of upper mantle, (This study) Paleocene Depth to Moho Paleocene-Eocene Miscellaneous : Permian Easting coordinate Pliocene Northing coordinate Pliocene-Quaternary Earthquakes spatial density*****, (This study) Precambrian Fault lines (Sahandi and Soheili 2014) density Quaternary Density of boundaries of geological units, (This study) * Minimum distance (calculated in this study) to the lithological units/ages (Sahandi and Soheili, 2014). **Averaged crustal magnetic susceptibility ( Teknik et al., 2020 ). *** The gravity gradients for the study area are at an elevation of 225 km above the Earth’s surface. The X-axis points to the north, the Y-axis points west, and the Z-axis points up ( Bouman et al., 2016 ). ****Vs indicate shear velocity ( Irandoust et al., 2022 ). ***** The location of seismic events (Mw ≥ 3; from 1940) ( http://www.isc.ac.uk/isc-ehb/search/catalogue ). 5. Methods In this study, ML-based methodologies were implemented on the diverse geospatial datasets for predicting metallogenic potential zones across the Iranian plateau. The workflow commenced with the compilation and harmonization of multi-source geospatial data, including lithological, structural, geophysical, and geochemical layers. These datasets were preprocessed to ensure spatial coordinate consistency. Then, each layer of the dataset is structured into a comprehensive geospatial database to be suitable for supervised learning. For the construction of a structured database, an initial feature selection process was conducted to identify the most relevant training features for metallogenic systems by employing the Minimum Redundancy and Maximum Relevance (MRMR) algorithm (Ding and Peng, 2005 ). This process utilized statistical analysis and regression-based filtering methods to reduce dimensionality and the effects of the noisy features, essential for improving the efficiency of subsequent modeling processes. A variety of machine learning algorithms were employed to determine the most effective method for predicting areas with high metallogenic potential. Hyperparameters have been optimized for each algorithm through grid-search methods and validation feedback to enhance predictive accuracy. Hyperparameters are typically configuration variables that should be regulated manually to manage machine learning model before training. The hyperparameters are coefficient or mathematical functions that manage the layer number and the size of a neural network, for instance. The hyperparameters of the algorithms (e.g., the learning rate and the batch size) optimize the model learning processes from the data. The trained model on the selected geospatial features, was subsequently applied across the study area, enabling the generation of comprehensive mineral prospectivity maps. In this study, we used three distinct datasets including a training dataset, 5-fold cross-validation, and an independent test set (Table S3). The training dataset consisted of 2037 spatial samples each associated with 69 geospatial features derived from geological, geophysical, and lithospheric features as listed in Table 2 and described in Section 4 . This dataset was used to fit the regression models. Model performance was validated by 5-fold cross-validation procedure on the training dataset. This technique involved partitioning the 2037 samples of the training dataset into five subsets. Four subsets were used iteratively for training and one was reserved for validation in each iteration, thereby ensuring robustness and reducing overfitting of the models. Model performance was averaged across all folds and evaluated using RMSE and R² metrics. Additionally, we have an independent test set of 70 well-documented major ore deposits (listed in Table 1 ), which was used to qualitatively assess the correctness of our models. The 70 major ore deposits were not involved in the training or validation processes and were only used to evaluate how well the final model predicted the location of known big metallic mineral. This separation helps us to check how well the model works both on the data it was trained with and on the coordinate of unseen data. Additionally, correlation and misfit metrics between the predicted and true occurrence of metallogenic intensity were conducted to assess model reliability. For the models with suboptimal performance (e.g., high RMS or low R 2 values), additional tuning was undertaken. This included removing outliers, refining hyperparameters, modifying feature selection criteria, and adjusting the optimization strategy to improve prediction accuracy and reduce model bias (Fig. 6 ). Seven distinct ML algorithms with their 24 subtypes (Table S1), available in the MATLAB regression learner toolbox, were selected to represent a diverse range of algorithm families including tree-based, kernel-based, probabilistic, and neural network methods. The aim is to select an algorithm with optimal performance, suitable for MPM in the Iranian plateau. These major algorithms include: (i) Linear Regression, (ii) Decision Tree, (iii) Vector Machine (SVM), (iv) Ensemble, (v) Gaussian Process Regression, (vi) Neural Network, and (vii) Kernel. Among these, Ensemble-based learning Methods, particularly those employing bagging (bootstrap aggregation), exhibited the most consistent and robust performance. Ensemble learning methods integrate predictions from multiple base models to improve the overall prediction of the model by enhancing model accuracy and reducing variance. In this study, the Bagging ensemble method uses several base models on bootstrapped subsets of the training data in parallel. The base models are primarily neural network models. The final output was derived by aggregating the most frequent or average individual predictions of the base models (Fig. 6 ). The design and execution of this methodology were successfully used in the recent prospecting and exploration studies (e.g., Yin and Li, 2022 ; Chen and Chen, 2023 ; He et al., 2024 ). The application of the Bagging Ensemble method in this study not only enhanced predictive stability but also showcased its effectiveness in integrating geospatial and geophysical data for mineral prospectivity mapping (MPM). Once the optimum model by tuning the hyperparameters is achieved, the trained model used to make predictions over regular grid nodes with a spatial resolution of 2 km × 2 km across the Iranian plateau. The cell size is selected based on the average resolution of the geophysical and geological raster grids. The gridded dataset has the same coverage as the aeromagnetic grid of Iran, which covers most of the plateau except for some gaps in the Zagros and Kopeh-Dagh regions (Teknik and Ghods, 2017 ; Teknik et al., 2020 ). We cast all datasets into a grid and name those as gridded datasets. It is important to note that this gridded dataset differs from the training feature dataset, which only includes sampling points located at known mineral occurrences. The training dataset consists of 2037 sampling points, while the gridded dataset consists of 377200 sampling points or grid nodes as summarized in Table S3. All the raster grids and rasterized grid layers were then resampled and aligned to the grid coordinate system, ensuring spatial and projection consistency with the training features and target variable, used for the final model prediction. 6. Results An ML-based approach was developed to predict the spatial variation of metallogenic intensity. To achieve this, we integrated the spatial information of Fe, Cu, Pb, Zn, Au, Ag, and Mg into the metallogenic intensity dataset serving as a training target (Fig. 2 b). This approach incorporated a comprehensive set of geophysical measurements and geological features (Table 2 ) as training datasets. The training datasets comprised 2,037 observation samples. Model performance was assessed through 5-fold cross-validation. Furthermore the validation is qualitatively assessed by using an independent observation point consisting of approximately 70 known metallic deposit locations. Model training was executed on MATLAB's machine learning toolbox using a system equipped with a 7-core Intel i7 GPU @ 2.8 GHz with 64 GB RAM. The training time is not equal for all models. The training models like Ensemble Bagged Trees and Neural Networks require more training time compared to simpler models like Linear Regression. The training was performed by using the parallel computing toolbox of MATLAB R2023b. The cross-validation and hyperparameter tuning were parallelized within loops and model-level parallel execution, which significantly reduced training time in the order of 3–5 hours, depending on hyperparameters variations. An analysis of the training features' importance was conducted to evaluate the contribution of each training feature. Features with high importance alongside qualitative geological studies can enhance the accuracy of the prospective mapping. Moreover, it provides insights for the development of targeted exploration plans in the future as well as deepens our understanding of regional-scale ore formation processes. The Minimum Redundancy and Maximum Relevance (MRMR) algorithm (Ding and Peng, 2005 ) was used for feature importance analysis. The MRMR algorithm assesses the importance of each training feature on the stable prediction of the model, which is spatial mineralization density in this study. It works by reducing the redundancy among the training features while ensuring their high relevance to the prediction parameter. To achieve this, the mutual information is used to evaluate the features interaction with each other and their relevance to the prediction parameter of the model. Through the importance evaluation process, it removes each training feature from the training process and then estimates the accuracy parameters (e.g., R 2 and RMS) of the trained model. Reduction of the prediction accuracy of the trained model indicates the high importance of the removed training feature. The distribution of importance of the training features shown in Fig. 7 indicates that geological (e.g., density of fault lines) and some geophysical (e.g., magnetic and its derivatives) features exhibit high importance. The lithospheric features (e.g., Moho and seismic velocity variations) together with gravity anomalies have lower importance, likely due to their low spatial resolution. Among the seven distinct ML algorithms with their 24 subtypes (Table S1), the Ensemble Bagged Trees algorithm demonstrated superior predictive accuracy, achieved by the highest coefficient of determination (R²) of approximately 0.98 and the lowest root mean square error (RMSE) of about 0.03 (Table 3 ). The estimated mineral prospectivity map (Fig. 8 ) illustrates the spatial density of the mineral occurrence across the Iranian plateau, achieved by Ensemble Bagged Trees algorithm. The high correlation between predicted and the observed values (R²≈0.98) underscores the robust performance of the supervised Ensemble Bagged Trees algorithm (Fig. 9 ). The Bagged Trees algorithm is very similar to the Random Forest algorithms. Random Forest algorithms are successfully used for MPM (Carranza and Laborte, 2015 ; Parsa and Maghsoudi, 2021 ; Rodriguez-Galiano, et al. 2015 ; Yang, et al 2022 ). Therefore, it is not a surprise that we got the highest performance using Bagged Trees methods (Table S1 in the supplementary). This regional-scale prospectivity map (Fig. 8 ) effectively identifies zones with known ore occurrences and predicts underexplored metallogenic zones that are primarily located in northwest, west, and Central Iran (Fig. 8 ). These prospective zones are predominantly associated with large-scale faults or major tectonic boundaries as well as magmatic complexes. Notably, the results indicate a southward extension of the high metallogenic potential of KaraDagh copper zone, in the northwest of the Iranian plateau. The southward extension is bounded by the NW-SE trending North Tabriz fault. Despite the limited number of training points, the model predicts spatial extension of the Malayer-Isfahan lead-zinc mineral zone toward the west, while a lower spatial density is anticipated eastward along the Kashan-Natanz metallogenic zone (Fig. 8 ). Additionally, areas nearby the central part of the Zagros suture zone located between Kermanshah and Neyriz ophiolites have been identified as potential targets for further exploration. To highlight the underexplored metallogenic zones, a residual density map (Fig. 10 ) was calculated by subtracting the observed spatial mineralization density (Fig. 2 b) from the predicted density of mineralization density (Fig. 8 ). This approach facilitates the identification of areas with high metallogenic potential that have not yet been thoroughly investigated. Interestingly, this residual prediction identifies three significant potential metallogenic zones, indicating new prospective areas in the northwest, western, and northern parts of the Iranian plateau. Table S1 The results are achieved for ML methods. RMSE is the root mean squared error on the validation set. The metrics of R-squared (R2) indicates how well the predicted results explain the target of training. MAE is the Mean absolute error. The MAE is always positive and similar to the RMSE, but less sensitive to outliers. MSE represents mean squared error. The list is sorted in ascending order of RMSE and descending order of R-squared values. Methods Type Subtype RMSE (Validation) MSE (Validation) RSquared (Validation) MAE (Validation) Ensemble Bagged Trees 0.03125 0.00100 0.97620 0.01809 Gaussian Process Regression Matem 5/2 GPR 0.03272 0.00107 0.97245 0.01818 Gaussian Process Regression Rational Quadratic GPR 0.03286 0.00108 0.97221 0.01821 Gaussian Process Regression Squared Exponential 0.03337 0.00111 0.97133 0.01850 Neural Network Bi-layered NN 0.03637 0.00132 0.96595 0.02410 Neural Network Tri-layered NN 0.03979 0.00158 0.95924 0.02600 Gaussian Process Regression Exponential 0.04054 0.00164 0.95770 0.02379 Neural Network Medium NN 0.04410 0.00194 0.94995 0.02773 SVM Cubic 0.04447 0.00198 0.94909 0.02742 Kernel SVM 0.04651 0.00216 0.94432 0.02977 Neural Network Wide NN 0.04832 0.00234 0.93989 0.02886 SVM Quadratic 0.04948 0.00245 0.93699 0.03293 Neural Network Narrow 0.05437 0.00296 0.92391 0.03477 SVM Medium Gaussian 0.05508 0.00303 0.92190 0.03575 Tree Fine Tree 0.05723 0.00328 0.91569 0.03054 Kernel Least Squares Regression kernel 0.05964 0.00356 0.90843 0.04204 Tree Medium 0.06426 0.00413 0.89369 0.03781 Ensemble Boosted Trees 0.06535 0.00427 0.89008 0.05057 Tree Coarse Tree 0.08223 0.00676 0.82594 0.05707 SVM Fine Gaussian 0.09723 0.00945 0.75665 0.05928 Linear Regression Linear 0.11999 0.01440 0.62938 0.09175 SVM Linear 0.15443 0.02385 0.38612 0.09373 SVM Coarse 0.17181 0.02952 0.24011 0.10043 Linear Regression Robust 0.17434 0.03039 0.21760 0.09108 . Table S2 Summary of ML Model Hyperparameters and Tuning Ranges Model Type Subtype Key Hyperparameters Tuning Ranges / Notes Ensemble (Bagged Trees) Bagged Trees NumLearningCycles, MinLeafSize NumLearningCycles: 30; MinLeafSize: 8 Gaussian Process Matern 5/2 Kernel KernelScale, Sigma, BaiscFunction, KernelScale: auto (fitrgp default), Sigma: auto, BasicFunction: constant Neural Network Tri-layered Feedforward NumHiddenLayers, NeuronsPerLayer, LearningRate Layers: 3 (fixed); Neurons: 10–100 per layer; LearningRate: 0.001–0.05 SVM Cubic BoxConstraint, KernelScale BoxConstraint: 0.1–100; KernelScale: 0.01–10 Kernel Regression SVM Kernel KernelFunction, Regularization, andIterationLimit KernelFunction: 'gaussian'; Regularization: 0.001–1; IterationLimit: 1000 Decision Tree Fine Tree Surrogate decision splits, MinLeafSize Surrogate decision splits: off; MinLeafSize : 1–5 (optimized at 4) Linear Regression Linear None (least-squares solution) Robust option: on Note: All the models were trained using MATLAB R2023b functions with custom tuning via grid search. The final parameters were selected using 5-fold cross-validation for minimizing RMSE. Table 3 The higher metrics results are achieved for the selected seven different ML methods. The metrics of root mean squared error (RMSE) and R-squared (R2) indicate how well the predicted results explain the target of training. The list is sorted in ascending order of RMSE and, simultaneously, in descending order of R-squared values. The training point number of observations is 2037 with 69 training features or predictors. Validation of the training results is tested with the 5-fold cross-validation. The optimum result is achieved with the Ensemble Bagged Trees method. For details of the training metrics results see Table S1. The best predicted spatial density of the ore occurrences across the model achieved by the Ensemble Bagged Trees method is shown in Fig. 8 . The predictive model for the six other models is presented in Figures S1 to S6 in the supplementary. Num. Method Sub Type RMSE (Validation) R_Squared (Validation) 1 Ensemble Bagged Trees 0.031 0.98 2 Gaussian Process Regression Matem 5/2 0.033 0.97 3 Neural Network Trilayered 0.040 0.96 4 SVM Cubic 0.044 0.95 5 kernel SVM 0.046 0.94 6 Tree Fine 0.057 0.91 7 Linear Regression Linear 0.119 0.62 Table S3 Summary of Data Flow Dataset Size Purpose Used For Training Dataset 2037 Model learning Regression modeling Validation (Cross validation folding) 5 subsets of training data Hyperparameter tuning & model selection RMSE/R² evaluation Independent Test Set 70 External model validation Final qualitatively asses of the trained model Gridded dataset 377200 Apply the trained model on the nodes of the grid Prediction map across the study area with a 2 km × 2 km. 7. Discussion 7.1. The mineral prospectivity mapping algorithmic perspective The data-driven approach of mineral prospectivity mapping (MPM) introduces regional-scale exploration solutions by integrating multidimensional geospatial dataset. This geospatial dataset typically consists of various geological features (e.g., rock units and fault lines), lithospheric parameters (e.g., Moho depth and seismic velocity heterogeneity of different layers of the lithosphere), and geophysical measurements (e.g., gravity, magnetic, and their respective derivatives). All of those heterogenous features merged into a unified training dataset. A set of supervised regression-based algorithms are used to derive a predictive model for delineating high-potential metallogenic zones across the Iranian plateau. Using seven major ML algorithms with their 24 subtypes of MATLAB regression learner toolbox, the Ensemble Bagged Trees algorithm performed more accurate and stable predictions for mineral prospectivity across the Iranian plateau, as evidenced with the lowest validation RMSE (~ 0.03), highest R² (~ 0.98) on other metrics (See details at Table S1 in the supplementary). The algorithm inherently has abilities of variance reduction, robustness to overfitting, and handling high-dimensional non-linear geospatial data (e.g., Yin and Li, 2022 ; Chen and Chen, 2023 ; He et al., 2024 ). Bagging or bootstrap aggregation lunches multiple decision trees using random subsets of the training data set and then aggregates their outputs to one solution. This approach reduces the potential high variance in the prediction of decision trees by averaging the predictions of all trees. Therefore, using this technique leads to a robust prediction model that is less sensitive to outlier prediction. In our study, the 69 training features represents complex and diverse geological, geophysical, and lithospheric ore-forming controlling parameters. This kind of multidimensional and often spatially noisy data can lead to overfitting in the individual learners. The Bagged Trees method enhances model stability by capturing the dominant prediction without being highly influenced by localized anomalies. The model's ability to derive the non-linear relationships between the training target of ore deposits spatial density and training features is essential for accurate prediction. The strong performance was observed in the cross-validation results. Moreover, the high spatial correlation of the independent 70 major deposit with areas of high values of the predicted density of mineralization, qualitatively provided extra assessment of the performance of the employed model. The Ensemble Bagged Trees algorithm shows the relatively accurate prediction of mineralization density, which is important for real-world mineral prospecting mapping. Importantly, this method aligned well with the goals of our study to detect underexplored metallogenic zones. The algorithm’s predictive strengths helped us to delineate new potential zones across a geologically complex and tectonically active region of Iranian plateau with diverse types of metallogenic zones, which made the prediction geologically interpretable and highly suitable for regional-scale mineral prospectivity mapping. 7.2. Limitations and challenges of ML-driven mineral prospective mapping Although the employed ML-driven prospective mapping enhances exploration efficiency in detecting high-potential metallogenic zones, it is essential to recognize its limitations to ensure accurate interpretation. The predictive accuracy of these models is fundamentally dependent on the quality, resolution, and temporal relevance of both training and target datasets (Farahnakian et al., 2024 ). For instance, the dataset of ore deposits used in this study is limited to the most recent (2017 and later) discovered ore deposits. Consequently, the results should be interpreted cautiously, with an awareness of potential gaps or outdated information that may affect the accuracy and comprehensiveness of the analysis (Singer, 2007 ; Mateus and Martins, 2019 ). It is crucial to integrate all deposit types as a unique training target data set; otherwise, prediction by individual AI predictive models would be impossible. By considering metallogenic intensity (spatial density of the ore deposits) parameter, we made a unified model of the diverse types of the seven selected ore deposits. Our approach treats the spatial density of all types of ore deposits as a proxy for metallogenic intensity parameter. This means that metallogenic intensity or spatial density of metallogeny across the plateau, structurally and lithologically associated with the same crustal-scale processes (e.g., Sun et al., 2020 ). While different ore deposit types (e.g., porphyry, skarn, vein, SEDEX, MVT) are genetically distinct, but they may still occur within shared mineralisation systems that are governed by common lithospheric-scale processes, including magmatism-related sources, faulting, fluid flow pathways, and depositional environments (McCuaig and Hronsky, 2017 ). Following the mineral systems approach (e.g., Hronsky and Groves, 2008 ; McCuaig et al., 2010 ), we argue that these different ore deposits can be regionally associated as an unique parameter, due to their formation within large-scale crustal architectures that condition the source, transport, and focusing of mineralizing fluids. By considering this argument, the similar evidences of different types of ore deposits can be tracked in the geological, geophysical and geochemical observations (Rodriguez-Galiano et al., 2015 ). Recent mineral system analysis is gradually accepting this approach in ore genesis studies. However, the conceptual framework of linking the mineral system to available data sets is still under development (McCuaig et al., 2010 ; Tagwai et al., 2024 ). While the mineral exploration studies have recognised the approach of this study as useful for the for prospecting metallogenic zones by investigation of ore formation processes (Knox-Robinson and Wyborn, 1997 ; Groves et al., 2022 ; Tagwai et al., 2024 ), using the spatial density as a proxy for mineralization intensity is not intended to imply genetic uniformity, but rather to model the cumulative metallogenic potential imparted by tectonomagmatic controls. This strategy enables us to capture regional mineralization patterns without violating the theoretical distinctions between deposit types. Our results support this view, as spatially clustered mineralization zones align with known trans-lithospheric faults, magmatic arcs, and high-relief tectonic boundaries—structural features widely recognized as metallogenic drivers. Despite these limitations, the MPM approach offers an organized approach that prioritizes data-driven targets, minimizes expenses, and enhances discovery rates. To prevent over-interpretation, it is essential to acknowledge the limitations of the input datasets. Integrating advanced ML techniques with high-quality updated geospatial data enables MPM to optimize exploration strategies and enhance the likelihood of successful ore deposit explorations (Zhang et al., 2024 ). A central challenge in this study arises from the irregular distribution of the discovered ore deposits (Fig. 2 ), which underscores both the geological complexity of the Iranian plateau and the influence of historically biased exploration efforts (Lou and Liu, 2023 ). The observed spatial pattern, marked by clustering and sparsity, highlights the challenges of directly attributing raster grid training features to discrete deposit locations due to the variable intensity of the mineralization. The irregular spatial distribution of the ore deposits can cause less uncertain prediction, especially where the distribution is sparse or different ore deposits are highly clustered. To overcome these limitations and align with advancements in handling imbalanced geospatial data, the spatial density of the mineral deposits is selected as the target variable for ML model, instead of depending on specific mineral deposit coordinates (Farahnakian et al., 2024 ). To evaluate the spatial uncertainty of the predictions, a residual map (Figure S7) was generated, showing the difference between true and predicted spatial densities of ore deposits at known locations. Residuals mostly fall within ± 0.1, suggesting good model performance overall. However, larger residuals (>|±0.1|) are observed in under-sampled areas, indicating greater model uncertainty. These residuals are particularly notable for isolated deposits distant from major metallogenic clusters. Interestingly, even in some highly sampled regions, such as the KaraDagh metallogenic zone in the northwest of the Iranian plateau, a localized cluster of elevated residuals is detected, possibly due to local geological complexity. Relatively low feature importance scores of lithospheric and gravity features (e.g., Moho, crustal seismic velocity, and Bouguer gravity anomalies) are in contrast with the generally accepted importance of lithospheric structures in controlling metallogenic zones. The Moho depth and Bouguer anomalies have spatially low resolution ( ~ > 0.5°). These low-resolution lithospheric features contribute less to localized variance of the metallogenic spatial density (Fig. 7 ). However, in this study qualitatively and less quantitatively, they still provide important geological context by delineating major crustal domains, tectonic boundaries, and zones of magmatic activities, all of which influence ore-forming processes. It is worth noting that the ore deposits dataset used here was last updated in 2017. Therefore, any exploration conducted since then could be vital in independently validating or refining our predictions. Therefore, newly identified ore occurrences would provide valuable ground-truth evidence for the model's reliability and help fine-tune prospectivity interpretations. 7.3. Geological implications of the results The suitable condition for the formation of metallic ore deposits is inherently associated with a specific physicochemical condition, geodynamic processes, and tectonic evolution. Therefore, the spatial mineral occurrences are typically attributed to the geological features (e.g., lithological units, fault systems, magmatic intrusions, and hydrothermal fluids). The geophysical anomalies (e.g., gravity and magnetic) reflect those geological structures, depending their sensitivity. These complex ore deposit formation mechanisms within various types of geological structures and their complicated geophysical signatures pose significant challenges to regional metallogenic zone mapping (Chernicoff et al., 2002 ; Bierlein et al., 2006 ; Leclerc et al., 2012 ; Dufréchou et al., 2015 ; Jamali and Mehrabi, 2015 ; Bauer et al., 2022 ). The residual spatial density of the mineral deposits (Fig. 10 ) was derived by subtracting the calculated spatial density of the ore deposits (Fig. 2 a) from the model’s prediction (Fig. 8 ). The residual density map provides insights to pinpoint areas across the Iranian plateau with high metallogenic potential, which have remained underexplored. The results indicate that the promising areas are closely associated with the known major metallogenic zones (Fig. 10 ). The residual anomalies are spatially aligned with established metallogenic provinces, such as the KaraDagh, Malayer-Isfahan, and Toroud zones (see Figs. 2 d and 10 ). This pattern indicates that the model accurately represents the main litho-structural and magmatic-tectonic controls on ore formation in the region. The zones lie within geodynamically active regions shaped by the convergence of the Arabian and Eurasian plates, where processes such as crustal shortening, strike-slip faulting, and arc magmatism have historically facilitated the formation of systems including porphyry Cu-Au-Mo ore deposits. The Central Iran block (e.g., Yazd, Tabas, and Lut blocks) exhibits comparatively weak residual signals, which presumably indicates that significant mineralized belts in this region have been well-mapped by previous explorations. However, some limited, structurally aligned anomalies trace the terrane boundaries and mapped faults. This observation highlights the significance of deep crustal fractures in directing hydrothermal fluids in central Iran. We argue that the distribution of ore deposits is not controlled only by active faults but controlled by all crustal scale fractures, old faults, and inactive lineaments. A better fault and lineament map will affect our results to be more precise but because of using a coarse fault map, the effect of all crustal-scale faults and fractures is missing in our results. However, the topography and magnetic field with their derivatives may provide the expected effect of those crustal-scale cracks, old faults, and inactive lineaments, which are not presented in the fault map of Iran. Especially, the high feature importance of the DEM (Digital Elevation Model) layer (Fig. 7 ) and its derivatives indicate its typical correspondence to the tectonically active regions such as fault zones, terrane boundaries, suture zones and magmatic intrusions, which are spatially associated with ore deposit (e.g., porphyry Cu, skarn, epithermal Au) (e.g., Yang et al., 2022 ; Zuo and Xu, 2023 ). Therefore, the high feature importance of topography does not imply that elevation causes mineralization, but instead the DEM acts as a proxy variable that captures the spatial pattern of geological processes, which are conducive to ore deposit formation. For instance, in the Iranian plateau, the Urumieh-Dokhtar magmatic arc, important for hosting porphyry and skarn deposits, is marked with high relief originating from intense subduction-related magmatism. The prospectivity maps also hold relevance for critical minerals of rare earth elements, which are often genetically related to metallic ore-forming systems (e.g., Petrella et al., 2014 ; Zarasvandi et al., 2015 ; Abedini et al., 2023 ). 7.4. AI-methods potentials for MPM and suggestions for future studies This study indicates the efficiency of applying modern ML algorithms on the various geospatial data for prospecting metallogenic zones on a regional scale, especially in areas with diverse tectonics, such as Iranian plateau. Employing this approach provided valuable regional insights into potential metallogenic zones by including only positive training dataset. Positive training data points indicate observation points with confirmed mineralization, while negative data points represent locations where ore deposit formation is unlikely. The applicability of the AI-driven approaches can be broadened through the integration of advanced deep learning methods by including both positive and negative training datasets. The aim is to provide interpretable AI-aided algorithms in mineral exploration with a promising trajectory in MPM. Enhancing the workflow by using new ML methods alongside advanced hyperparameter optimization would help mitigate overfitting and more effectively manage non-linearity problems between training dataset and the target of the method (Chen and Wu, 2017 ; Li et al., 2021 , 2024 ; Maepa et al., 2021 ; Parsa and Carranza, 2021 ; Parsa and Maghsoudi, 2021 ; N. Yang et al., 2022 ; Yin et al., 2023 ; Zuo and Xu, 2023 ; Wake et al., 2024 ). To achieve a trained model with higher prediction accuracy, it is essential to incorporate diverse geospatial datasets with higher resolution. For instance, integration of radiometric and geochemical sampling together with satellite-driven multispectral or hyperspectral imagery can significantly improve the prediction accuracy of the trained model. These additional datasets, especially hyperspectral images, identify hydrothermal alteration indicators, geochemical anomalies, and structurally controlled mineralizing systems. Applying this approach will enhance the model sensitivity to characteristics such as iron zones, clay alteration, or silicification, which typically indicative of deposit types like porphyries, epithermal systems, or skarns. While this study focused on identifying prospective metallogenic zones, regardless of mineralization type, the workflow could be adapted to distinguish deposit types by classification learning algorithms using spectral, geophysical, or geochemical training dataset (Abedi et al., 2012 ; Shirmard et al., 2022 ; Chen and Chen, 2023 ; Farahnakian et al., 2024 ; Mahboob et al., 2024 ). For example, segmentation algorithms could relatively differentiate porphyry copper systems, characterized by halos and disseminated sulfides alteration, from epithermal gold-silver deposits, which are marked by alteration zones of vein-hosted metallogenic and intense silicification (Shayeganpour and Tangestani, 2022 ; Amraoui et al., 2025 ; Farahbakhsh et al., 2025 ). Employing such approaches would facilitate more precise targeting of specific deposit types. These approaches, when properly implemented, would yield significant insights into spatial patterns of the ore deposits, thereby enhancing mineral exploration strategies and decision-making processes. 8. Conclusions The AI-aided mineral prospectivity mapping (MPM) approach is used to identify new metallogenic zones in the Iranian plateau. Analyzing regional-scale geological and geophysical data is typically associated with challenges when using traditional methods. AI algorithms facilitate resolving the nonlinear and complex patterns among irregularly distributed mineral deposits using multimodal geological and geophysical big data. Regression learning methods are used to identify the intrinsic relationship between various geological and geophysical training features and the locations of the metallic ore deposits as training targets. The training target dataset is formed by integrating the location of seven metallic ore deposits including Fe, Cu, Pb, Zn, Au, Ag, and Mg. Then the spatial density of ore deposits is computed using the integrated dataset. The information regarding the grade and tonnage of various ore deposits in our dataset is limited, thus identical importance weights are assigned to each ore deposit. To form training feature dataset, the vector-based features (e.g., lithological units and fault lines) were transformed into a set of raster grids to ensure their data format consistency with geophysical and lithospheric raster grids. The raster grids and vector-based features are created by calculating the minimum distance to geological units and lithospheric structural boundaries. These created raster grids together with raster grids of geophysical measurements and mid-lithospheric interfaces (e.g., average crustal shear waves and Moho depth), formed the training dataset comprising 69 features. Feature importance analysis is conducted to identify the most relevant features for ore formation. The learning performance is evaluated for seven major regressions-based algorithms with their subtypes reaching 24 methods. The Ensemble Bagged Trees method displayed the best performance. The selection criteria were the minimum RMSE (~ 0.03) and maximum R-Squared (R 2 ≈ 0.98) value from the regression analysis of prediction versus observation of validation data. Once the model is trained and its accuracy adheres to the acceptable critical, it is employed to estimate the spatial density of the ore occurrences at grid nodes with a cell spacing of 2 km across the Iranian plateau. The predicted spatial density of the ore deposits is compared with locations of comprehensive data on world-class ore deposits. To highlight underexplored areas, residual spatial density anomalies were calculated by subtracting the observed from the predicted spatial density. The results indicate the prediction is optimally aligned with the spatial association of major ore deposits. The trends of predicted spatial density indicates a new zone of metallogenic activity close to the existing metallogenic zones. This study suggests that the structural boundaries and magmatic complexes along with geophysical observations, are significant indicators of metallic ore deposits and should be prioritized in future mineral exploration plans. The results indicate that the high-potential metallogenic zones are primarily located in the northwest of the study area, aligning closely with the known distribution of the Copper metallogenic zone of the KaraDagh. The southern boundary of this zone is primarily outlined by the major North Tabriz fault. The predicted area on the western side of the Zagros suture aligns closely with the known Malayer-Isfahan Pb-Zn metallogenic zone. The residual spatial density of the ore deposits suggests that Central Iran has already been almost fully explored, revealing fewer unexplored metallogenic zones. Nonetheless, a locally less explored metallogenic zone has been identified near the Bafgh, Nehbandan-Ferdous, and Jiroft-Shaherbabak metallogenic zones. The results of this study are useful for enhancing local-scale exploration success by employing a practical and multidisciplinary workflow of reginal-scale mineral prospective mapping (MPM). This approach addresses the rising demands for ore minerals, considering sustainable development and the economic limitations of mineral exploration. Leveraging the recent advancements in the big data analysis and artificial intelligence algorithms aid environmentally sustainable mineral exploration. Declarations Data/Software availability statement The National Centers for Environmental Information ( https://www.ncei.noaa.gov/products/etopo-global-relief-model ) was used to access the ETOPO1 global elevation model (Amante and Eakins, 2009 ) from the ETOPO Global Relief Model grid. The Cenozoic volcano’s locations are downloaded from the Smithsonian Institution's Global Volcanism Program (GVP) website ( https://volcano.si.edu/ ). The gravity gradient grids at 225 km and 255 km height are available from the website of the European Space Agency (ESA) ( https://earth.esa.int/eogateway/catalog/goce-global-gravity-field-models-and-grids ) (Bouman et al., 2016 ). The EGM2008 global Bouguer anomaly model (Pavlis et al., 2012) has been accessed by the National Geospatial-Intelligence Agency from the website ( https://bgi.obs-mip.fr/grids-and-models-2/grids-and-models-2-2/#toc7 ). The MATLAB codes are used to prepare data and compute the models that can be provided by the author. The lithological units/ages (Sahandi and Soheili 2014). Averaged crustal magnetic susceptibility (Teknik et al. 2020 ). Vs indicate shear velocity (Irandoust et al., 2022 ). The location of seismic events (Mw ≥ 3; from 1940) ( http://www.isc.ac.uk/isc-ehb/search/catalogue ). Conflict of Interest The authors have no conflicts of interest to declare that are relevant to the content of this article Acknowledgments This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors. Vahid Teknik (VT) extends his special appreciation to Ebrahim Gholamzadeh, whose support made this research possible. VT thanks Irina Artemieva and Hans Thybo for helpful scientific support. Amr Abdelnasser extends his appreciation to the Scientific Research Project (BAP Project ID: 46804) at Istanbul Technical University (ITU, Turkey). References Abedi M, Norouzi GH, Bahroudi A (2012) Support vector machine for multi-classification of mineral prospectivity areas. Comput Geosci 46:272–283. https://doi.org/10.1016/j.cageo.2011.12.014 Abedini M, Ziaii M, Timkin T, Pour AB (2023) Machine Learning (ML)-Based Copper Mineralization Prospectivity Mapping (MPM) Using Mining Geochemistry Method and Remote Sensing Satellite Data. Remote Sens 2023 15(15):3708. https://doi.org/10.3390/RS15153708 Adiri Z, Lhissou R, El Harti A, Jellouli A, Chakouri M (2020) Recent advances in the use of public domain satellite imagery for mineral exploration: A review of Landsat-8 and Sentinel-2 applications. Ore Geol Rev 117. https://doi.org/10.1016/j.oregeorev.2020.103332 Aghazadeh M, Hou Z, Badrzadeh Z, Zhou L (2015) Temporal-spatial distribution and tectonic setting of porphyry copper deposits in Iran: Constraints from zircon U-Pb and molybdenite Re-Os geochronology. Ore Geol Rev 70:385–406. https://doi.org/10.1016/J.OREGEOREV.2015.03.003 Alaminia Z, Karimpour MH, Homam SM, Finger F (2013) The magmatic record in the Arghash region (northeast Iran) and tectonic implications. Int J Earth Sci 102:1603–1625. https://doi.org/10.1007/s00531-013-0897-1 Alavi M (1996) Tectonostratigraphic synthesis and structural style of the Alborz Mountain System in Iran. J Geodyn 21:1–33 Aliyari F, Rastad E, Mohajjel M (2012) Gold Deposits in the Sanandaj-Sirjan Zone: Orogenic Gold Deposits or Intrusion-Related. Gold Systems? Resource Geol 62:296–315. https://doi.org/10.1111/J.1751-3928.2012.00196 .X;PAGEGROUP:STRING:PUBLICATION Allen MB, Kheirkhah M, Neill I, Emami MH, Mcleod CL (2013) Generation of Arc and Within-plate Chemical Signatures in Collision Zone Magmatism: Quaternary Lavas from Kurdistan Province, Iran. J Petrol 54:887–911. https://doi.org/10.1093/petrology/egs090 Amante C, Eakins BW (2009) ETOPO1 1 Arc-Minute Global Relief Model: Procedures, Data Sources and Analysis, NOAA Technical Memorandum NESDIS NGDC-24. https://doi.org/10.1594/PANGAEA.769615 Amraoui T, Ibouh H, Farah A, Bammou Y, Shebl A (2025) Remote sensing mapping of structural and hydrothermal alteration in the mougueur inlier, Eastern high atlas, Morocco. Sci Rep 15:1–18. https://doi.org/10.1038/S41598-025-99402-0 ;SUBJMETA=213,2151,330,431,704;KWRD=GEOLOGY,MINERALOGY,PETROLOGY Anderson ED, Monecke T, Hitzman MW, Zhou W, Bedrosian PA (2017) Mineral Potential mapping in an accreted island-Arc setting using aeromagnetic data: An example from Southwest Alaska. Econ Geol 112:375–396. https://doi.org/10.2113/ECONGEO.112.2.375 Arjmandzadeh R, Karimpour MH, Mazaheri SA, Santos JF, Medina JM, Homam SM (2011) Sr-Nd isotope geochemistry and petrogenesis of the Chah-Shaljami granitoids (Lut Block, Eastern Iran). J Asian Earth Sci 41:283–296. https://doi.org/10.1016/J.JSEAES.2011.02.014 Asadi S, Moore F, Zarasvandi A (2014) Discriminating productive and barren porphyry copper deposits in the southeastern part of the central Iranian volcano-plutonic belt, Kerman region, Iran: A review. Earth Sci Rev 138:25–46. https://doi.org/10.1016/J.EARSCIREV.2014.08.001 Bagheri H (2015) Crustal lineament control on mineralization in the Anarak area of Central Iran. Ore Geol Rev 66:293–308. https://doi.org/10.1016/J.OREGEOREV.2014.10.028 Bauer TE, Lynch EP, Sarlus Z, Drejing-Carroll D, Martinsson O, Metzger N, Wanhainen C (2022) Structural Controls on Iron Oxide Copper-Gold Mineralization and Related Alteration in a Paleoproterozoic Supracrustal Belt: Insights from the Nautanen Deformation Zone and Surroundings, Northern Sweden. Econ Geol 117:327–359. https://doi.org/10.5382/ECONGEO.4862 Benaissi L, Tarek A, Tobi A, Ibouh H, Zaid K, Elamari K, Hibti M (2022) Geological mapping and mining prospecting in the Aouli inlier (Eastern Meseta, Morocco) based on remote sensing and geographic information systems (GIS). China Geol 5:614–625. https://doi.org/10.31035/cg2022035 Berberian M, King GCP (1981) Towards a paleogeography and tectonic evolution of Iran. Can J Earth Sci 18:210–265. https://doi.org/10.1139/e81-019 Bergen KJ, Johnson PA, de Hoop MV, Beroza GC (2019) Machine learning for data-driven discovery in solid Earth geoscience. Science (1979) 363. https://doi.org/10.1126/science.aau0323 Beygi S, Talovina IV, Tadayon M, Pour AB (2021) Alteration and structural features mapping in Kacho-Mesqal zone, Central Iran using ASTER remote sensing data for porphyry copper exploration. Int J Image Data Fusion 12:155–175. https://doi.org/10.1080/19479832.2020.1838628 Bierlein FP, Groves DI, Goldfarb RJ, Dubé B (2006) Lithospheric controls on the formation of provinces hosting giant orogenic gold deposits. Min Depos 40:874–886. https://doi.org/10.1007/S00126-005-0046-2 Boulin J (1991) Structures in Southwest Asia and evolution of the eastern Tethys. Tectonophysics 196:211–268. https://doi.org/10.1016/0040-1951(91)90325-M Bouman J, Ebbing J, Fuchs M, Sebera J, Lieb V, Szwillus W, Haagmans R, Novak P (2016) Satellite gravity gradient grids for geophysics. Sci Rep 6:1–11. https://doi.org/10.1038/srep21050 Bruzzone L, Chi M, Marconcini M (2006) A novel transductive SVM for semisupervised classification of remote-sensing images. IEEE Trans Geosci Remote Sens 44:3363–3373. https://doi.org/10.1109/tgrs.2006.877950 Burg JP (2018) Geology of the onshore Makran accretionary wedge: Synthesis and tectonic interpretation. Earth Sci Rev 185:1210–1231. https://doi.org/10.1016/j.earscirev.2018.09.011 Carranza EJM (2009) Controls on mineral deposit occurrence inferred from analysis of their spatial pattern and spatial association with geological features. Ore Geol Rev 35:383–400. https://doi.org/10.1016/j.oregeorev.2009.01.001 Carranza EJM, Laborte AG (2015) Random forest predictive modeling of mineral prospectivity with small number of prospects and data with missing values in Abra (Philippines). Comput Geosci 74:60–70. https://doi.org/10.1016/j.cageo.2014.10.004 Chen G, Cheng Q, Zuo R, Liu T, Xi Y (2015) Identifying gravity anomalies caused by granitic intrusions in Nanling mineral district, China: a multifractal perspective. Geophys Prospect 63:256–270. https://doi.org/10.1111/1365-2478.12187 Chen J, Chen Y (2023) A high-performance voting-based ensemble model of graph convolutional extreme learning machines for identifying geochemical anomalies related to mineralization. Ore Geol Rev 162:1–14. https://doi.org/10.1016/j.oregeorev.2023.105706 Chen Y, Wu W (2017) Mapping mineral prospectivity using an extreme learning machine regression. Ore Geol Rev 80:200–213. https://doi.org/10.1016/j.oregeorev.2016.06.033 Chernicoff CJ, Richards JP, Zappettini EO (2002) Crustal lineament control on magmatism and mineralization in northwestern Argentina: Geological, geophysical, and remote sensing evidence. Ore Geol Rev 21:127–155. https://doi.org/10.1016/S0169-1368(02)00087-2 Chiu HY, Chung SL, Zarrinkoub MH, Mohammadi SS, Khatib MM, Iizuka Y (2013) Zircon U-Pb age constraints from Iran on the magmatic evolution related to Neotethyan subduction and Zagros orogeny. Lithos 162–163:70–87. https://doi.org/10.1016/J.LITHOS.2013.01.006 Chukwu C, Betts P, Moore D, Munukutla R, Armit R, McLean M, Grose L (2024) Unsupervised machine learning and depth clusters of Euler deconvolution of magnetic data: a new approach to imaging geological structures. Explor Geophys 55:223–245. https://doi.org/10.1080/08123985.2023.2299475 Daliran F (2008) The carbonate rock-hosted epithermal gold deposit of Agdarreh, Takab geothermal field, NW Iran - Hydrothermal alteration and mineralisation. Min Depos 43:383–404. https://doi.org/10.1007/S00126-007-0167-X Davies RS, Davies MJ, Groves D, Davids K, Brymer E, Trench A, Sykes JP, Dentith M (2021) Learning and Expertise in Mineral Exploration Decision-Making: An Ecological Dynamics Perspective. Int J Environ Res Public Health 18:9752. https://doi.org/10.3390/ijerph18189752 Dentith M, Enkin RJ, Morris W, Adams C, Bourne B (2020) Petrophysics and mineral exploration: a workflow for data analysis and a new interpretation framework. Geophys Prospect 68:178–199. https://doi.org/10.1111/1365-2478.12882 Dentith M, Mudge S (2014) Geophysics for the mineral exploration geoscientist. AusIMM Bull. https://doi.org/10.1017/CBO9781139024358 Ding C, Peng H (2005) Minimum redundancy feature selection from microarray gene expression data. J Bioinform Comput Biol 3:185–205 Dufréchou G, Harris LB, Corriveau L, Antonoff V (2015) Regional and local controls on mineralization and pluton emplacement in the Bondy gneiss complex, Grenville Province, Canada interpreted from aeromagnetic and gravity data. J Appl Geophy 116:192–205. https://doi.org/10.1016/J.JAPPGEO.2015.03.015 Enayat M, Ghods A (2023) 3D Shear-Wave Velocity Model of Central Makran Using Ambient-Noise Adjoint Tomography. J Geophys Res Solid Earth 128. https://doi.org/10.1029/2023JB026928 ;PAGEGROUP:STRING:PUBLICATION e2023JB026928 Farahbakhsh E, Goel D, Pimparkar D, Muller RD, Chandra R (2025) Convolutional neural networks for mineral prospecting through alteration mapping with remote sensing data. https://doi.org/10.1007/s41064-025-00344-z Farahnakian F, Sheikh J, Zelioli L, Nidhi D, Seppä I, Ilo R, Nevalainen P, Heikkonen J (2024) Addressing imbalanced data for machine learning based mineral prospectivity mapping. Ore Geol Rev 174:106270. https://doi.org/10.1016/J.OREGEOREV.2024.106270 Geranian H, Tabatabaei SH, Asadi HH, Carranza EJM (2016) Application of Discriminant Analysis and Support Vector Machine in Mapping Gold Potential Areas for Further Drilling in the Sari-Gunay Gold Deposit, NW Iran. Nat Resour Res 25:145–159. https://doi.org/10.1007/S11053-015-9271-2 Golmohammadi A, Karimpour MH, Shafaroudi AM, Mazaheri SA (2015) Alteration-mineralization, and radiometric ages of the source pluton at the Sangan iron skarn deposit, northeastern Iran. Ore Geol Rev 65:545–563 Ghorbani M (2013a) The economic geology of Iran: Mineral deposits and natural resources. The Economic Geology of Iran: Mineral Deposits and Natural Resources 1–569. https://doi.org/10.1007/978-94-007-5625-0 Ghorbani M (2013b) Metallogenic and mining provinces, belts and zones of Iran. Springer Geol 199–295. https://doi.org/10.1007/978-94-007-5625-0_6/FIGURES/26 Glennie KW, Hughes Clarke MW, Boeuf MGA, Pilaar WFH, Reinhardt BM (1990) Inter-relationship of Makran-Oman Mountains belts of convergence. Geological Society, London. https://doi.org/10.1144/GSL.SP.1992.049.01.47 ., Special Publications Gonzalez-Alvarez I, Goncalves MA, Carranza EJM (2020) Introduction to the Special Issue Challenges for mineral exploration in the 21st century: Targeting mineral deposits under cover. Ore Geol Rev 126:103785. https://doi.org/10.1016/j.oregeorev.2020.103785 Groves DI, Santosh M, Müller D, Zhang L, Deng J, Yang LQ, Wang QF, Mineral systems: Their advantages in terms of developing holistic genetic models and for target generation in global mineral exploration. Geosystems and Geoenvironment 1., Bierlein DI (2022) F.P., 2007. Geodynamic settings of mineral deposit systems. J Geol Soc London 164, 19–30. https://doi.org/10.1144/0016-76492006-065 Guo P, Yang T (2023) Quantifying Continental Crust Thickness Using the Machine Learning Method. J Geophys Res Solid Earth 128:1–16. https://doi.org/10.1029/2022JB025970 Hajsadeghi S, Mirmohammadi M, Asghari O, Meshkani SA (2018) Geology and mineralization at the copper-rich volcanogenic massive sulfide deposit in Nohkouhi, Posht-e-Badam block, Central Iran. Ore Geol Rev 92:379–396. https://doi.org/10.1016/J.OREGEOREV.2017.11.030 Hassanlouei BT, Rajabzadeh MA (2019) Iron ore deposits associated with Hormuz evaporitic series in Hormuz and Pohl salt diapirs, Hormuzgan province, southern Iran. J Asian Earth Sci 172:30–55. https://doi.org/10.1016/j.jseaes.2018.08.024 Hassanpour S, Rajabpour S (2020) Magmatic-hydrothermal evolution of the Anjerd Cu skarn deposit, NW Iran: perspectives on mineral chemistry, fluid inclusions and stable isotopes. Ore Geol Rev 117:103269 He H, Zhu H, Yang X, Zhang W, Wang J (2024) Mineral prospectivity prediction based on convolutional neural network and ensemble learning. Scientific Reports 2024 14:1 14, 1–21. https://doi.org/10.1038/s41598-024-73357-0 Hezarkhani A (2006) Petrology of the intrusive rocks within the Sungun porphyry copper deposit, Azerbaijan, Iran. J Asian Earth Sci 27(3):326–340 Hezarkhani A (2008) Hydrothermal evolution of the Miduk Porphyry copper system, Kerman, Iran: A fluid inclusion investigation. Int Geol Rev 50:665–684. https://doi.org/10.2747/0020-6814.50.7.665 Holden EJ, Dentith M, Kovesi P (2008) Towards the automated analysis of regional aeromagnetic data to identify regions prospective for gold deposits. Comput Geosci 34:1505–1513. https://doi.org/10.1016/J.CAGEO.2007.08.007 Hronsky JMA, Groves DI (2008) Science of targeting: Definition, strategies, targeting and performance measurement. Aust J Earth Sci 55:3–12. https://doi.org/10.1080/08120090701581356;WGROUP:STRING:PUBLICATION Irandoust MA, Priestley K, Sobouti F (2022) High-Resolution Lithospheric Structure of the Zagros Collision Zone and Iranian Plateau. J Geophys Res Solid Earth 127. https://doi.org/10.1029/2022JB025009 Jamali H, Mehrabi B (2015) Relationships between arc maturity and Cu-Mo-Au porphyry and related epithermal mineralization at the Cenozoic Arasbaran magmatic belt. Ore Geol Rev 65:487–501. https://doi.org/10.1016/J.OREGEOREV.2014.06.017 Kesler SE, Simon AC (2015) Mineral Resources, Economics and the Environment. https://doi.org/10.7302/22482 Kheyrollahi H, Alinia F, Ghods A (2018) Regional magnetic lithologies and structures as controls on porphyry copper deposits: Evidence from Iran. Explor Geophys 49:98–110. https://doi.org/10.1071/EG16042 Knox-Robinson CM, Wyborn LAI (1997) Towards a holistic exploration strategy: Using Geographic Information Systems as a tool to enhance exploration. Aust J Earth Sci 44:453–463. https://doi.org/10.1080/08120099708728326 Korehie MT, Ardebili O, Ahari HD, Shirzad MR, Fotovati V, Ghalamghash J, Kiani T, Najafi A, Soltani NS, Ashtiani ME, Farhatjah B (2019) Atlas of Iran’s Geology and Mineral Distribution. Springer, Springer Nature Leclerc F, Harris LB, Bédard JH, Van Breemen O, Goulet N (2012) Structural and stratigraphic controls on magmatic, volcanogenic, and shear zone-hosted mineralization in the Chapais-Chibougamau mining camp, northeastern Abitibi, Canada (1,2). Econ Geol 107:963–969. https://doi.org/10.2113/ECONGEO.107.5.963 Li Q, Chen G, Wang D (2024) Mineral Prospectivity Mapping Using Semi-supervised Machine Learning. https://doi.org/10.1007/S11004-024-10161-6/FIGURES/16 . Math Geosci 1–31 Li S, Chen J, Liu C, Wang Y (2021) Mineral prospectivity prediction via convolutional neural networks based on geological big data. J Earth Sci 32:327–347. https://doi.org/10.1007/s12583-020-1365-z Li YE, O’Malley D, Beroza G, Curtis A, Johnson P (2023) Machine Learning Developments and Applications in Solid-Earth Geosciences: Fad or Future? J Geophys Res Solid Earth 128. https://doi.org/10.1029/2022JB026310 Lösing M, Ebbing J (2021) Predicting Geothermal Heat Flow in Antarctica With a Machine Learning Approach. J Geophys Res Solid Earth 126:1–16. https://doi.org/10.1029/2020JB021499 Lou Y, Liu Y (2023) Mineral Prospectivity Mapping of Tungsten Polymetallic Deposits Using Machine Learning Algorithms and Comparison of Their Performance in the Gannan Region, China. Earth Space Sci 10. https://doi.org/10.1029/2022EA002596 . e2022EA002596 Maepa F, Smith RS, Tessema A (2021) Support vector machine and artificial neural network modelling of orogenic gold prospectivity mapping in the Swayze greenstone belt, Ontario, Canada. Ore Geol Rev 130:103968. https://doi.org/10.1016/j.oregeorev.2020.103968 Maghfouri S, Hosseinzadeh MR (2018) The early Cretaceous Mansourabad shale-carbonate hosted Zn-Pb (-Ag) deposit, central Iran: An example of vent-proximal sub-seafloor replacement SEDEX mineralization. Ore Geol Rev 95:20–39 Maghfouri S, Rastad E, Momenzadeh M, Movahednia M, Hashempour SS, Jadehkenary KA, Ghaderi M, Konari MB, Izanloo J, Yarmohammadi A, Peernajmodin H (2025) Malayer-Esfahan Metallogenic Belt, Iran: Jurassic-Early Cretaceous sediment-(volcanic) hosted Zn-Pb (± Ba ± Ag), Fe-Mn-Pb (± Ba ± Cu) and Ba (± Pb ± Zn) deposits with evolution of basin. Journal of the Geological Society, jgs2024-284. Mahboob MA, Celik T, Genc B (2024) Predictive modelling of mineral prospectivity using satellite remote sensing and machine learning algorithms. Remote Sens Appl 36:101316. https://doi.org/10.1016/J.RSASE.2024.101316 Mateus A, Martins L (2019) Challenges and opportunities for a successful mining industry in the future. Boletín Geológico y Min 130:99–121. https://doi.org/10.21701/BOLGEOMIN.130.1.007 McCuaig TC, Beresford S, Hronsky J (2010) Translating the mineral systems approach into an effective exploration targeting system. Ore Geol Rev 38:128–138. https://doi.org/10.1016/J.OREGEOREV.2010.05.008 McCuaig TC, Hronsky J (2017) The mineral systems concept: the key to exploration targeting. Appl Earth Sci 126:77–78. https://doi.org/10.1080/03717453.2017.1306274 McMillan M, Haber E, Peters B, Fohring J (2021) Mineral prospectivity mapping using a VNet convolutional neural network. Lead Edge 40:99–105. https://doi.org/10.1190/tle40020099.1 Mehrdar A, Motaghi K, Ghods A, Sobouti F, Priestley K, Pachhai S, Shabanian E, Zarunizadeh Z, Zeynaddini-Meymand R, El-Hussain I (2025) Crustal and uppermost mantle structure of the Iranian Makran subduction zone from ambient noise and earthquake surface wave tomography. Geophys J Int 241:70–85. https://doi.org/10.1093/GJI/GGAE419 Meshkani SA, Mehrabi B, Yaghubpur A, Sadeghi M (2013) Recognition of the regional lineaments of Iran: Using geospatial data and their implications for exploration of metallic ore deposits. Ore Geol Rev 55:48–63. https://doi.org/10.1016/J.OREGEOREV.2013.04.007 Mehrabi B, Siani MG, Goldfarb R, Azizi H, Ganerod M, Marsh EE (2016) Mineral assemblages, fluid evolution, and genesis of polymetallic epithermal veins, Glojeh district, NW Iran. Ore Geol Rev 78:41–57 Mollai H, Sharma R, Pe-Piper G (2009) Copper mineralization around the Ahar batholith, north of Ahar (NW Iran): Evidence for fluid evolution and the origin of the skarn ore deposit. Ore Geol Rev 35:401–414. https://doi.org/10.1016/j.oregeorev.2009.02.005 Monsef I, Monsef R, Mata J, Zhang Z, Pirouz M, Rezaeian M, Esmaeili R, Xiao W (2018) Evidence for an early-MORB to fore-arc evolution within the Zagros suture zone: Constraints from zircon U-Pb geochronology and geochemistry of the Neyriz ophiolite (South Iran). Gondwana Res 62:287–305. https://doi.org/10.1016/J.GR.2018.03.002 Monsef I, Rahgoshay M, Pirouz M, Chiaradia M, Grégoire M, Ceuleneer G (2019) The Eastern Makran Ophiolite (SE Iran): evidence for a Late Cretaceous fore-arc oceanic crust. Int Geol Rev 61:1313–1339. https://doi.org/10.1080/00206814.2018.1507764 Monsef I, Zhang Z, Shabanian E, le Roux P, Rahgoshay M (2022) Tethyan subduction and Cretaceous rift magmatism at the southern margin of Eurasia: Evidence for crustal evolution of the South Caspian Basin. Earth Sci Rev 228:104012. https://doi.org/10.1016/J.EARSCIREV.2022.104012 Moritz R, Ghazban F, Singer BS (2006) Eocene gold ore formation at Muteh, Sanandaj-Sirjan tectonic zone, western Iran: A result of late-stage extension and exhumation of metamorphic basement rocks within the Zagros orogen. Econ Geol 101:1497–1524. https://doi.org/10.2113/GSECONGEO.101.8.1497 Mousivand F, Rastad E, Peter JM, Maghfouri S (2018) Metallogeny of volcanogenic massive sulfide deposits of Iran. Ore Geol Rev 95:974–1007 Nabatian G, Rastad E, Neubauer F, Honarmand M, Ghaderi M (2015) Iron and Fe–Mn mineralisation in Iran: implications for Tethyan metallogeny. Aust J Earth Sci 62:211–241. https://doi.org/10.1080/08120099.2015.1002001 Omidianfar S, Monsef I, Rahgoshay M, Zheng J, Cousens B (2020) The middle Eocene high-K magmatism in Eastern Iran Magmatic Belt: constraints from U-Pb zircon geochronology and Sr-Nd isotopic ratios. Int Geol Rev 62:1751–1768. https://doi.org/10.1080/00206814.2020.1716272 Parsa M, Carranza EJM (2021) Modulating the impacts of stochastic uncertainties linked to deposit locations in data-driven predictive mapping of mineral prospectivity. Nat Resour Res 30:3081–3097. https://doi.org/10.1007/s11053-021-09891-9 Parsa M, Maghsoudi A (2021) Assessing the effects of mineral systems-derived exploration targeting criteria for random Forests-based predictive mapping of mineral prospectivity in Ahar-Arasbaran area, Iran. Ore Geol Rev 138:104399. https://doi.org/10.1016/j.oregeorev.2021.104399 Pavlis NK, Holmes SA, Kenyon SC, Factor JK (2012a) The development and evaluation of the Earth Gravitational Model 2008 (EGM2008). J Geophys Res Solid Earth 117:1–38. https://doi.org/10.1029/2011JB008916 Pavlis NK, Holmes SA, Kenyon SC, Factor JK (2012b) The development and evaluation of the Earth Gravitational Model 2008 (EGM2008). J Geophys Res Solid Earth 117:1–38. https://doi.org/10.1029/2011JB008916 Penney C, Tavakoli F, Saadat A, Nankali HR, Sedighi M, Khorrami F, Sobouti F, Rafi Z, Copley A, Jackson J, Priestley K (2017) Megathrust and accretionary wedge properties and behaviour in the Makran subduction zone. Geophys J Int 209:1800–1830. https://doi.org/10.1093/gji/ggx126 Petrella L, Williams-Jones AE, Goutier J, Walsh J (2014) The nature and origin of the rare earth element mineralization in the misery syenitic intrusion, Northern Quebec, Canada. Econ Geol 109:1643–1666. https://doi.org/10.2113/ECONGEO.109.6.1643 Qaderi S, Maghsoudi A, Pour AB, Yousefi M (2024) Geological Controlling Factors on Mississippi Valley-Type Pb-Zn Mineralization in Western Semnan, Iran. Minerals (2075-163X), 14(9). Rajabi A, Rastad E, Canet C (2013) Metallogeny of Permian-Triassic carbonate-hosted Zn-Pb and F deposits of Iran: A review for future mineral exploration. Aust J Earth Sci 60(2):197–216 Richards JP (2015) Tectonic, magmatic, and metallogenic evolution of the Tethyan orogen: From subduction to collision. Ore Geol Rev 70:323–345. https://doi.org/10.1016/J.OREGEOREV.2014.11.009 Richards JP, Wilkinson D, Ullrich T (2006) Geology of the Sari Gunay epithermal gold deposit, northwest Iran. Econ Geol 101:1455–1496. https://doi.org/10.2113/GSECONGEO.101.8.1455 Rodriguez-Galiano V, Sanchez-Castillo M, Chica-Olmo M, Chica-Rivas M, Machine learning predictive models for mineral prospectivity: An evaluation of neural networks, random forest, regression trees and support vector machines. Ore Geol Rev 71, 804–818., Sahandi M, Soheili M (2015) 2014. Geological Map of Iran, scale 1:1000000. Geological Survey of Iran Shahabpour J (1999) The role of deep structures in the distribution of some major ore deposits in Iran, NE of the Zagros thrust zone. J Geodyn 28:237–250. https://doi.org/10.1016/S0264-3707(98)00040-4 Shayeganpour S, Tangestani MH (2022) Extraction of rock and alteration geons by FODPSO segmentation and GP regression on the HyMap imagery: A case study of SW Birjand, Eastern Iran. Ore Geol Rev 143:104767. https://doi.org/10.1016/J.OREGEOREV.2022.104767 Shirmard H, Farahbakhsh E, Müller RD, Chandra R (2022) A review of machine learning in processing remote sensing data for mineral exploration. Remote Sens Environ 268. https://doi.org/10.1016/j.rse.2021.112750 Singer DA (2007) Estimating amounts of undiscovered mineral resources. In Proceedings for a Workshop on Deposit Modeling, Mineral Resource Assessment, and Their Role in Sustainable Development: USGS Circular. (Vol. 1294, pp. 79–84) Soloviev SG, Kryazhev SG, Dvurechenskaya SS, Vasyukov VE, Shumilin DA, Voskresensky KI (2019) The superlarge Malmyzh porphyry Cu-Au deposit, Sikhote-Alin, eastern Russia: Igneous geochemistry, hydrothermal alteration, mineralization, and fluid inclusion characteristics. Ore Geol Rev 113. https://doi.org/10.1016/j.oregeorev.2019.103112 Stampfli GM (2000) Tethyan oceans. Geological Society, London, Special Publications 173, 1–23. https://doi.org/10.1144/GSL.SP.2000.173.01.01 Stampfli GM, Borel GD (2004) The TRANSMED Transects in Space and Time: Constraints on the Paleotectonic Evolution of the Mediterranean Domain. Springe, Berlin, pp 53–90 Stampfli GM, Borel GD (2002) A plate tectonic model for the Paleozoic and Mesozoic constrained by dynamic plate boundaries and restored synthetic oceanic isochrons. Earth Planet Sci Lett 196:17–33. https://doi.org/10.1016/S0012-821X(01)00588-X Stöcklin J (1974) Possible Ancient Continental Margins in Iran. The Geology of Continental Margins. Springer, Berlin Heidelberg, pp 873–887. https://doi.org/10.1007/978-3-662-01141-6_64 Stocklin J (1968) Structural History and Tectonics of Iran: A Review. Am Assoc Pet Geol Bull 52:1229–1258 Stöcklin J (1968) Structural history and tectonics of Iran: a review. Am Assoc Pet Geol Bull 52:1229–1258 Stosch HG, Romer RL, Daliran F, Rhede D (2011) Uranium-lead ages of apatite from iron oxide ores of the Bafq District, East-Central Iran. Min Depos 46:9–21. https://doi.org/10.1007/S00126-010-0309-4 Sun T, Li H, Wu K, Chen F, Zhu Z, Hu Z (2020) Data-driven predictive modelling of mineral prospectivity using machine learning and deep learning methods: a case study from southern Jiangxi Province China. Minerals 10:102. https://doi.org/10.3390/min10020102 Tagwai MG, Jimoh OA, Shehu SA, Zabidi H (2024) Application of GIS and remote sensing in mineral exploration: current and future perspectives. World J Eng 21:487–502. https://doi.org/10.1108/WJE-09-2022-0395 TaleFazel E, Mehrabi B, GhasemiSiani M (2019) Epithermal systems of the Torud-Chah Shirin district, northern Iran: Ore-fluid evolution and geodynamic setting. Ore Geol Rev 109:253–275 Teknik V (2024) The improved Moho depth imaging in the Arabia-Eurasia collision zone: A machine learning approach integrating seismic observations and satellite gravity data. Tectonophysics 893:230553. https://doi.org/10.1016/j.tecto.2024.230553 Teknik V, Artemieva IM, Thybo H (2024) Limited arc magmatism and seismicity due to extensive mantle wedge serpentinization in the Makran subduction zone. Earth Planet Sci Lett 645:118950. https://doi.org/10.1016/j.epsl.2024.118950 Teknik V, Ghods A (2017) Depth of magnetic basement in Iran based on fractal spectral analysis of aeromagnetic data. Geophys J Int 209:1878–1891. https://doi.org/10.1093/gji/ggx132 Teknik V, Thybo H, Artemieva IM, Ghods A (2020) A new tectonic map of the Iranian plateau based on aeromagnetic identification of magmatic arcs and ophiolite belts. Tectonophysics 792. https://doi.org/10.1016/J.TECTO.2020.228588 Torab FM, Lehmann B (2007) Magnetite-apatite deposits of the Bafq district, Central Iran: apatite geochemistry and monazite geochronology. Mineral Mag 71:347–363. https://doi.org/10.1180/MINMAG.2007.071.3.347 Verdel C, Wernicke BP, Hassanzadeh J, Guest B (2011) A Paleogene extensional arc flare - up in Iran. Tectonics 30:1–20. https://doi.org/10.1029/2010TC002809 Wake N, Farahbakhsh E, Müller RD (2024) Lateritic Ni–Co Prospectivity Modeling in Eastern Australia Using an Enhanced Generative Adversarial Network and Positive-Unlabeled Bagging. https://doi.org/10.1007/S11053-024-10423-4 . Natural Resources Research Woodhead J, Landry M (2021) Harnessing the Power of Artificial Intelligence and Machine Learning in Mineral Exploration—Opportunities and Cautionary Notes. SEG Discovery 19–31. https://doi.org/10.5382/Geo-and-Mining-13 Xiong Y, Zuo R (2020) Recognizing multivariate geochemical anomalies for mineral exploration by combining deep learning and one-class support vector machine. Comput Geosci 140. https://doi.org/10.1016/j.cageo.2020.104484 Xiong Y, Zuo R (2018) GIS-based rare events logistic regression for mineral prospectivity mapping. Comput Geosci 111:18–25. https://doi.org/10.1016/j.cageo.2017.10.005 Xiong Y, Zuo R, Carranza EJM (2018) Mapping mineral prospectivity through big data analytics and a deep learning algorithm. Ore Geol Rev 102:811–817. https://doi.org/10.1016/j.oregeorev.2018.10.006 Yang H, huan, Wang Q, Li Ybo, Lin B, Song Y, Wang Y, yun, He W, Li H, wei, Li S, Li J, li, Liu C, cheng, Feng S, bin, Xin T, Fu Xlian, Liang X, juan, Zhang Q, Wang Bqi, Li Y (2022) Geology and mineralization of the Tiegelongnan supergiant porphyry-epithermal Cu (Au, Ag) deposit (10 Mt) in western Tibet, China: A review. China Geol 5:136–159. https://doi.org/10.1016/S2096-5192(22)00091-X Yang N, Zhang Z, Yang J, Hong Z (2022) Mineral prospectivity prediction by integration of convolutional autoencoder network and random forest. Nat Resour Res 31:1103–1119. https://doi.org/10.1007/s11053-022-10038-7 Yin B, Zuo R, Sun S (2023) Mineral prospectivity mapping using deep self-attention model. Nat Resour Res 32:37–56. https://doi.org/10.1007/s11053-022-10142-8 Yin J, Li N (2022) Ensemble learning models with a Bayesian optimization algorithm for mineral prospectivity mapping. Ore Geol Rev 145. https://doi.org/10.1016/j.oregeorev.2022.104916 Zarasvandi A, Liaghat S, Zentilli M (2005) Geology of the Darreh-Zerreshk and Ali-Abad porphyry copper deposits, central Iran. Int Geol Rev 47(6):620–646 Zarasvandi A, Liaghat S, Zentilli M, Reynolds PH (2007) 40Ar/39Ar geochronology of alteration and petrogenesis of porphyry copper-related granitoids in the Darreh-Zerreshk and Ali-Abad area, central Iran. Explor Min Geol 16:11–24. https://doi.org/10.2113/GSEMG.16.1-2.11 Zarasvandi A, Rezaei M, Sadeghi M, Lentz D, Adelpour M, Pourkaseb H (2015) Rare earth element signatures of economic and sub-economic porphyry copper systems in Urumieh-Dokhtar Magmatic Arc (UDMA), Iran. Ore Geol Rev 70:407–423. https://doi.org/10.1016/J.OREGEOREV.2015.01.010 Zarasvandi A, Rezaei M, Raith JG, Asadi S, Lentz D (2019) Hydrothermal fluid evolution in collisional Miocene porphyry copper deposits in Iran: Insights into factors controlling metal fertility. Ore Geol Rev 105:183–200 Zhang SE, Lawley CJM, Bourdeau JE, Nwaila GT, Ghorbani Y (2024) Nat Resour Res 2024 33(3 33):995–1023. https://doi.org/10.1007/S11053-024-10322-8 . Workflow-Induced Uncertainty in Data-Driven Mineral Prospectivity Mapping Zhao P (2007) Quantitative mineral prediction and deep mineral exploration. Earth Sci Front 14:1–10 Zuo R, Xu Y (2023) Graph deep learning model for mapping mineral prospectivity. Math Geosci 55:1–21. https://doi.org/10.1007/s11004-022-10015-z Zuo R, Peng Y, Li T, Xiong Y (2021) Challenges of geological prospecting big data mining and integration using deep learning algorithms. Earth Sci 46:350–358 Additional Declarations The authors declare no competing interests. Supplementary Files Supplementary.docx Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-7730637","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":521613090,"identity":"de21326c-f58f-4c34-94c1-aa0c97e2ffed","order_by":0,"name":"Vahid Teknik","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA8ElEQVRIie3PuwrCMBSA4RMCdfGypij6CimdBMVXsXRwqc5uFgJ18bIqir6Fs8HBRXcho2uHuol0MG0dXGzrJph/SDKcjyQAKtVPVoR9UW4EEINAHrRCPkJjghYRwTkIJAQwjq6DLEKPZ773IazrK8bM1mPXKGNAwc1JIadBl6+BmtUaZ3Z/JgwPA9aXu89Edx16kA+zNsRih/5EIEk0XEojcz8mo5g0J6KTSSokuaVblcSGu7ByEJ/yNTWN5cJixtQVtocRS/2LVnHMwB/WG+TSu5J7KNrbMePBLYUk0deOvHh1s+bfC78ZVqlUqn/pCbEqTFg4YObBAAAAAElFTkSuQmCC","orcid":"","institution":"","correspondingAuthor":true,"prefix":"","firstName":"Vahid","middleName":"","lastName":"Teknik","suffix":""},{"id":521613091,"identity":"1b65ab61-aee3-4dbc-a3f6-694b31004ca6","order_by":1,"name":"Iman Monsef","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Iman","middleName":"","lastName":"Monsef","suffix":""},{"id":521613092,"identity":"90d4bf93-3519-49ba-8100-64f8e7777d30","order_by":2,"name":"Amr Abdelnasser","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Amr","middleName":"","lastName":"Abdelnasser","suffix":""},{"id":521613093,"identity":"7560ef20-976b-4899-8b06-0595f46e2758","order_by":3,"name":"Abdolreza Ghods","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Abdolreza","middleName":"","lastName":"Ghods","suffix":""}],"badges":[],"createdAt":"2025-09-27 22:35:06","currentVersionCode":1,"declarations":{"humanSubjects":false,"vertebrateSubjects":false,"conflictsOfInterestStatement":false,"humanSubjectEthicalGuidelines":false,"humanSubjectConsent":false,"humanSubjectClinicalTrial":false,"humanSubjectCaseReport":false,"vertebrateSubjectEthicalGuidelines":false},"doi":"10.21203/rs.3.rs-7730637/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7730637/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":92485017,"identity":"c348e223-360e-4de1-adeb-34d593c923bf","added_by":"auto","created_at":"2025-09-30 08:25:49","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":13870377,"visible":true,"origin":"","legend":"","description":"","filename":"MineralMLv20250815.docx","url":"https://assets-eu.researchsquare.com/files/rs-7730637/v1/326e88e60c98b11f20e95085.docx"},{"id":92485225,"identity":"374239c9-c06a-474c-b20c-a505f19d484c","added_by":"auto","created_at":"2025-09-30 08:33:48","extension":"json","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":342,"visible":true,"origin":"","legend":"","description":"","filename":"rs7730637.json","url":"https://assets-eu.researchsquare.com/files/rs-7730637/v1/2671236bf33a270a0f809353.json"},{"id":92484991,"identity":"a5fca7cf-01cc-4fe7-9345-52ac2ee48958","added_by":"auto","created_at":"2025-09-30 08:25:48","extension":"xml","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":364398,"visible":true,"origin":"","legend":"","description":"","filename":"rs77306370enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-7730637/v1/80c19ce19b3f63b083d1df56.xml"},{"id":92485227,"identity":"7ae3a826-cb91-46a8-8e5c-788be799dc8f","added_by":"auto","created_at":"2025-09-30 08:33:48","extension":"jpeg","order_by":3,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":715569,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage1.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7730637/v1/232dd824c580b028ba544415.jpeg"},{"id":92485229,"identity":"b60f5dea-ff78-4557-804b-cd0fddbecfd2","added_by":"auto","created_at":"2025-09-30 08:33:48","extension":"jpeg","order_by":4,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":658367,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage10.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7730637/v1/12212730185ce92c9fdec731.jpeg"},{"id":92484115,"identity":"c2c5c58b-da43-4d1c-90ae-32b26204b1bc","added_by":"auto","created_at":"2025-09-30 08:17:48","extension":"jpeg","order_by":5,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1556721,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage11.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7730637/v1/468610d68723dc8fd4fefe3c.jpeg"},{"id":92484124,"identity":"aa801b8d-5257-448e-b6b3-78d0e454dd3f","added_by":"auto","created_at":"2025-09-30 08:17:48","extension":"jpeg","order_by":6,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":573736,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage12.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7730637/v1/6b26700b1b5025b3a2206f2f.jpeg"},{"id":92484119,"identity":"c525ed3a-a5de-4172-bf32-d24cbee640be","added_by":"auto","created_at":"2025-09-30 08:17:48","extension":"jpeg","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1579315,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage13.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7730637/v1/70a4d43b536aa774f3a5c07c.jpeg"},{"id":92484125,"identity":"3300d081-c8af-4b63-9536-0090549335c4","added_by":"auto","created_at":"2025-09-30 08:17:48","extension":"jpeg","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":555208,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage14.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7730637/v1/f294e86b4c215a1be3f104cc.jpeg"},{"id":92484998,"identity":"55a79a41-6869-4598-aff0-0921f75b6c1c","added_by":"auto","created_at":"2025-09-30 08:25:48","extension":"jpeg","order_by":9,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1639560,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage15.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7730637/v1/7c166c989a33928852ac70ae.jpeg"},{"id":92484997,"identity":"5261be9c-3da8-4941-831b-99abcd710394","added_by":"auto","created_at":"2025-09-30 08:25:48","extension":"jpeg","order_by":10,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":543444,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage16.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7730637/v1/9b44f26ddd2e45b0fabc19af.jpeg"},{"id":92484127,"identity":"dedd8d82-6a91-4eb6-a39d-0474fb6e15fb","added_by":"auto","created_at":"2025-09-30 08:17:48","extension":"jpeg","order_by":11,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1726364,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage17.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7730637/v1/dd090a6da21b01589d212fd2.jpeg"},{"id":92484131,"identity":"fa60b068-7d4a-478e-84a2-703a74f1826f","added_by":"auto","created_at":"2025-09-30 08:17:48","extension":"jpeg","order_by":12,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":589496,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage18.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7730637/v1/15aecd4d6b3ae65b70b0b4c4.jpeg"},{"id":92484132,"identity":"0bfeed41-e2a6-41f2-bcc9-27e0101aadaa","added_by":"auto","created_at":"2025-09-30 08:17:48","extension":"jpeg","order_by":13,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1556360,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage19.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7730637/v1/4cebcb99904983de4a1de341.jpeg"},{"id":92485231,"identity":"13ba1dd2-7d76-4f75-9791-08f7b8fbb40a","added_by":"auto","created_at":"2025-09-30 08:33:48","extension":"jpeg","order_by":14,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":769466,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage2.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7730637/v1/30c9ca3faeb99a396aa75d5f.jpeg"},{"id":92485003,"identity":"ae4b85a2-7a99-4849-bc9b-8aafc736031b","added_by":"auto","created_at":"2025-09-30 08:25:48","extension":"jpeg","order_by":15,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":615268,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage20.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7730637/v1/b4daedf9ccb7168f598b19f7.jpeg"},{"id":92485232,"identity":"61eeefe7-f5ea-46ff-8dba-eae8dd4fc169","added_by":"auto","created_at":"2025-09-30 08:33:48","extension":"jpeg","order_by":16,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1634897,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage21.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7730637/v1/5efd7afb2942d1aa3ff05647.jpeg"},{"id":92484134,"identity":"4dbe28de-2c65-498c-9c1f-bdc99233432f","added_by":"auto","created_at":"2025-09-30 08:17:48","extension":"jpeg","order_by":17,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":562244,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage22.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7730637/v1/363e836b11b267756441d6e4.jpeg"},{"id":92485000,"identity":"60760a73-deaf-49d2-99dd-49422e91fd19","added_by":"auto","created_at":"2025-09-30 08:25:48","extension":"png","order_by":18,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1016901,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage23.png","url":"https://assets-eu.researchsquare.com/files/rs-7730637/v1/165ace6780c1d122b4770124.png"},{"id":92485233,"identity":"10e13636-0e81-4552-9311-1da3670bc5ad","added_by":"auto","created_at":"2025-09-30 08:33:48","extension":"jpeg","order_by":19,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":904034,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage3.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7730637/v1/94ec5a7c7d878a65b54da44a.jpeg"},{"id":92485001,"identity":"8fd47674-d33d-4f14-bed6-59ee06885498","added_by":"auto","created_at":"2025-09-30 08:25:48","extension":"jpeg","order_by":20,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":595832,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage4.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7730637/v1/f8a8e2bdf5ce1c454f8a791d.jpeg"},{"id":92484137,"identity":"e2353426-6d15-4a4c-908b-dfa82df85315","added_by":"auto","created_at":"2025-09-30 08:17:48","extension":"jpeg","order_by":21,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":973682,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage5.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7730637/v1/67940269236abd0866399994.jpeg"},{"id":92484142,"identity":"90be12ef-48c8-4faa-9b85-3868c728a2d4","added_by":"auto","created_at":"2025-09-30 08:17:49","extension":"jpeg","order_by":22,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":286170,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage6.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7730637/v1/b1bbbbbbdd118e921b236f68.jpeg"},{"id":92484143,"identity":"6bb69314-d22b-440b-becb-824267cb2492","added_by":"auto","created_at":"2025-09-30 08:17:49","extension":"jpeg","order_by":23,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":343640,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage7.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7730637/v1/52ce1f4a3959f0c70b07b3f5.jpeg"},{"id":92484139,"identity":"60e8340d-2ae6-4ebe-9b27-7efe3b1b08a3","added_by":"auto","created_at":"2025-09-30 08:17:48","extension":"jpeg","order_by":24,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":615303,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage8.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7730637/v1/c777db783f661c78c7dae905.jpeg"},{"id":92484138,"identity":"c40cf774-29e9-4b6d-95e4-1a35ab007f1e","added_by":"auto","created_at":"2025-09-30 08:17:48","extension":"jpeg","order_by":25,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":230802,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage9.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7730637/v1/7edc49f4d0d5d1dc069e1ee6.jpeg"},{"id":92484148,"identity":"55ed5ef0-0247-42fb-8c89-592c7273c876","added_by":"auto","created_at":"2025-09-30 08:17:49","extension":"png","order_by":26,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":250170,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-7730637/v1/b41d76d6ce30f997878e2546.png"},{"id":92486398,"identity":"a7a4bd66-3703-4667-a07b-3a5708f0abd5","added_by":"auto","created_at":"2025-09-30 08:41:49","extension":"png","order_by":27,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":190102,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage10.png","url":"https://assets-eu.researchsquare.com/files/rs-7730637/v1/a34242e034f2fb6321fa25f9.png"},{"id":92485010,"identity":"03bf918b-de40-48e8-b625-13804e7c2181","added_by":"auto","created_at":"2025-09-30 08:25:49","extension":"png","order_by":28,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":372269,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage11.png","url":"https://assets-eu.researchsquare.com/files/rs-7730637/v1/a3ad8bd79ec0cffe8bb04694.png"},{"id":92485007,"identity":"99188d29-d3a8-416d-8b6c-1e77966de575","added_by":"auto","created_at":"2025-09-30 08:25:49","extension":"png","order_by":29,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":140430,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage12.png","url":"https://assets-eu.researchsquare.com/files/rs-7730637/v1/aa6c4eed4142f07135ad7da2.png"},{"id":92485004,"identity":"8c92b67a-c751-45ed-ac68-dd58f853b73c","added_by":"auto","created_at":"2025-09-30 08:25:49","extension":"png","order_by":30,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":379012,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage13.png","url":"https://assets-eu.researchsquare.com/files/rs-7730637/v1/75940135ea9e45b8f990816a.png"},{"id":92485012,"identity":"f60cfcbd-2c97-4ecf-a730-7291cf1ca77a","added_by":"auto","created_at":"2025-09-30 08:25:49","extension":"png","order_by":31,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":140541,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage14.png","url":"https://assets-eu.researchsquare.com/files/rs-7730637/v1/0af699b1b4fb6fdc07e2a04f.png"},{"id":92484141,"identity":"02a39100-c8e6-4279-b119-2008a4917317","added_by":"auto","created_at":"2025-09-30 08:17:49","extension":"png","order_by":32,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":397888,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage15.png","url":"https://assets-eu.researchsquare.com/files/rs-7730637/v1/9e2637e5b300d435eb1e2551.png"},{"id":92484157,"identity":"1949f152-2249-4e1a-a5c7-0c7d2d77ffdc","added_by":"auto","created_at":"2025-09-30 08:17:49","extension":"png","order_by":33,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":134530,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage16.png","url":"https://assets-eu.researchsquare.com/files/rs-7730637/v1/d853faf0b8ffa2c687882005.png"},{"id":92485234,"identity":"0c7c6feb-0543-4c99-8c51-33b601724a86","added_by":"auto","created_at":"2025-09-30 08:33:49","extension":"png","order_by":34,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":402770,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage17.png","url":"https://assets-eu.researchsquare.com/files/rs-7730637/v1/0190d716d4e9dca1b7a11838.png"},{"id":92484167,"identity":"c6bee93f-8c18-4087-9656-568148d10742","added_by":"auto","created_at":"2025-09-30 08:17:49","extension":"png","order_by":35,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":148493,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage18.png","url":"https://assets-eu.researchsquare.com/files/rs-7730637/v1/815896a80df1c90359c5ea51.png"},{"id":92484150,"identity":"128f2fe8-7ab1-41de-a11a-ba25606856ce","added_by":"auto","created_at":"2025-09-30 08:17:49","extension":"png","order_by":36,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":384110,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage19.png","url":"https://assets-eu.researchsquare.com/files/rs-7730637/v1/1e40c25c273c0d2e1567393d.png"},{"id":92484145,"identity":"bbbbbc33-a340-4d96-95ed-e52e8fc00778","added_by":"auto","created_at":"2025-09-30 08:17:49","extension":"png","order_by":37,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":280642,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-7730637/v1/24e7b4fc4aba4b7c89ebcb70.png"},{"id":92484159,"identity":"a3723b09-a665-4014-b2e6-3d5781fdc93a","added_by":"auto","created_at":"2025-09-30 08:17:49","extension":"png","order_by":38,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":154483,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage20.png","url":"https://assets-eu.researchsquare.com/files/rs-7730637/v1/7278a11e4483f99786928dce.png"},{"id":92485236,"identity":"98680850-1c65-41d2-a0d8-6472516e28c7","added_by":"auto","created_at":"2025-09-30 08:33:49","extension":"png","order_by":39,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":393383,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage21.png","url":"https://assets-eu.researchsquare.com/files/rs-7730637/v1/438d957b25d48ad52b4feeb2.png"},{"id":92484146,"identity":"932ab934-8ee9-4b2e-84d1-692a715b0b4f","added_by":"auto","created_at":"2025-09-30 08:17:49","extension":"png","order_by":40,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":203322,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage22.png","url":"https://assets-eu.researchsquare.com/files/rs-7730637/v1/bf498385f12ea6aab9d23463.png"},{"id":92485239,"identity":"2df198de-f10c-47f2-be98-794f7949d889","added_by":"auto","created_at":"2025-09-30 08:33:49","extension":"png","order_by":41,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":145113,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage23.png","url":"https://assets-eu.researchsquare.com/files/rs-7730637/v1/9a5b58d816e747120e4020ff.png"},{"id":92485019,"identity":"dbfe36d7-da34-4948-806e-e5b8340d69f4","added_by":"auto","created_at":"2025-09-30 08:25:49","extension":"png","order_by":42,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":473145,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-7730637/v1/f6e4e2e19269d78676796162.png"},{"id":92484151,"identity":"4a2f2921-4a8f-402c-a777-25d47b68f22c","added_by":"auto","created_at":"2025-09-30 08:17:49","extension":"png","order_by":43,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":243566,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-7730637/v1/e9b903d034cb564096a467cb.png"},{"id":92485235,"identity":"2a1c390e-0cdd-4bb2-8ce6-7bc66fd244cd","added_by":"auto","created_at":"2025-09-30 08:33:49","extension":"png","order_by":44,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":333569,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-7730637/v1/b3668436426c6cf0eb13d14e.png"},{"id":92484153,"identity":"b5747693-1685-467d-8cd8-5dd9b7bfc4a7","added_by":"auto","created_at":"2025-09-30 08:17:49","extension":"png","order_by":45,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":82736,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-7730637/v1/4c938f5054432b84f8b9083e.png"},{"id":92485238,"identity":"d3ae8723-3683-47f4-97ad-4bd7f49e3003","added_by":"auto","created_at":"2025-09-30 08:33:49","extension":"png","order_by":46,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":76780,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage7.png","url":"https://assets-eu.researchsquare.com/files/rs-7730637/v1/792aa9ec6e1df60aaca12f1d.png"},{"id":92485014,"identity":"b7be373d-133f-4a8d-8ee4-063c5ed5cbc6","added_by":"auto","created_at":"2025-09-30 08:25:49","extension":"png","order_by":47,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":215956,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage8.png","url":"https://assets-eu.researchsquare.com/files/rs-7730637/v1/4fb2bb0466d62b99310775c7.png"},{"id":92485008,"identity":"45141ae8-838c-498e-918a-44e8cf17a44d","added_by":"auto","created_at":"2025-09-30 08:25:49","extension":"png","order_by":48,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":90875,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage9.png","url":"https://assets-eu.researchsquare.com/files/rs-7730637/v1/b486813d81a4c1618c1a28eb.png"},{"id":92485009,"identity":"c9ec37f8-1444-484d-8c88-086d6d55cf0e","added_by":"auto","created_at":"2025-09-30 08:25:49","extension":"xml","order_by":49,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":362694,"visible":true,"origin":"","legend":"","description":"","filename":"rs77306370structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-7730637/v1/71f676fa7db22aef55471d3e.xml"},{"id":92484163,"identity":"f3b7a00d-a8e4-495f-8378-e6eef971de43","added_by":"auto","created_at":"2025-09-30 08:17:49","extension":"html","order_by":50,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":371351,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-7730637/v1/86441f2171589f7f39e57ebd.html"},{"id":92484987,"identity":"d6508d3b-0d74-4bbb-ba78-b79dcbede36b","added_by":"auto","created_at":"2025-09-30 08:25:48","extension":"jpeg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":715569,"visible":true,"origin":"","legend":"\u003cp\u003eSimplified Tectono-magmatic map of the Iranian plateau (modified after Sahandi and Soheili, 2014) overlaid by the location of the major Fe, Cu, Pb, Zn, and Au ore deposits (see Table 1; after \u0026nbsp;Meshkani et al., 2013). Abbreviations: DF, Dorouneh Fault; MRF, Main Recent Fault; MZT, Main Zagros Thrust; UDMA, Urumieh-Dokhtar magmatic arc. The Zagros suture zone lies approximately along MRF and MZT.\u003c/p\u003e","description":"","filename":"image1.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7730637/v1/dc69015b4b434efcf6159581.jpeg"},{"id":92484107,"identity":"144b2b48-b43b-4432-a86a-acd0a1380a72","added_by":"auto","created_at":"2025-09-30 08:17:48","extension":"jpeg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":769466,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003ea\u003c/strong\u003e) Distribution map of the seven ore deposits (Cu, Fe, Au, Pb-Zn, Ag, and Mg). The crosses indicate the location of the mineral deposits without detailed information on their grade and size of the deposits (Korehie et al., 2019). This dataset includes 2037 observation points used for training and validation with 5-fold cross-validation. The black triangles represent quaternary volcanic cones. The symbols other than crosses and black triangles are major known deposits (see Table 1; after \u0026nbsp;Meshkani et al., 2013), which are used for testing. The colour of the symbols indicates the type of mineral. Abbreviations: SA, Sarcheshmeh copper deposit; ME, Meiduk copper deposit; MA, Mazraeh copper deposit; SU, Sungon copper deposits \u003cstrong\u003eb\u003c/strong\u003e) The spatial density of the mineral occurrence with no detailed information. The pale crosses indicate the location of mineral deposits. \u003cstrong\u003ec\u003c/strong\u003e) The resampled spatial density of the mineral deposits at their corresponding locations. \u003cstrong\u003ed\u003c/strong\u003e) Metallogenic zones modified after Ghorbani (2013)and updated based on the high spatial density (\u0026gt; 0.2 num/sq km) of mineral deposits. The numbers correspond to the metallogenic zone names as follows: (1) KaraDagh; (2) Takab-Tarom-Hashtjin; (3) Malayer-Isfahan; (4) Kashan-Natanz; (5) Toroud; (6) Anarak; (7) QaleBala; (8) Bafgh; (9) Khaf; (10) Nehbandan-Ferdous; (11) Jiroft-Shahrebabak; (12) Kahnuj-Fanuj.\u003c/p\u003e","description":"","filename":"image2.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7730637/v1/afd9f89cb7f0f2a20635d38a.jpeg"},{"id":92485226,"identity":"52236c4b-d084-4c59-b8a6-1d0910709ee9","added_by":"auto","created_at":"2025-09-30 08:33:48","extension":"jpeg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":904034,"visible":true,"origin":"","legend":"\u003cp\u003eSelected geophysical training measurements, including \u003cstrong\u003ea\u003c/strong\u003e) digital elevation model from the ETOPO1 global elevation model (Amante and Eakins, 2009). The symbols indicated major known ore deposits (Table 1), which are grouped for quality checking of the final AI-model; \u003cstrong\u003eb\u003c/strong\u003e) the variable reduced-to-pole (VRTP) total magnetic intensity (TMI) (Teknik and Ghods, 2017). \u003cstrong\u003ec\u003c/strong\u003e) The Bouguer anomaly map derived by a compilation of terrestrial, altimetry-derived, and airborne gravity measurements, as incorporated in the EGM2008 global Bouguer anomaly model (Pavlis et al., 2012); and \u003cstrong\u003ed\u003c/strong\u003e) The vertical Vzz component of gravity gradients at 225 km above the Earth’s surface with respect to WGS84 (Bouman et al., 2016). To make the vertical derivative map (panel d), the Z-axis is considered to point up.\u003c/p\u003e","description":"","filename":"image3.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7730637/v1/880bd07ff075739f09bd40ea.jpeg"},{"id":92484111,"identity":"046421d5-e6f7-4bed-845f-94854092d28a","added_by":"auto","created_at":"2025-09-30 08:17:48","extension":"jpeg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":595832,"visible":true,"origin":"","legend":"\u003cp\u003eThe selected features of the calculated minimum distance to \u003cstrong\u003ea\u003c/strong\u003e) the faults (Sahandi and Soheili, 2014); \u003cstrong\u003eb\u003c/strong\u003e) the igneous outcrops; \u003cstrong\u003ec\u003c/strong\u003e) the ophiolite outcrops; \u003cstrong\u003ed\u003c/strong\u003e) the andesite outcrops;\u003cstrong\u003e e\u003c/strong\u003e) the Eocene igneous outcrops; and \u003cstrong\u003ef\u003c/strong\u003e) the granite outcrops.\u003c/p\u003e","description":"","filename":"image4.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7730637/v1/911fb031022dcd689a132175.jpeg"},{"id":92484113,"identity":"74709bf5-7ef0-4afd-a634-85f92b6aea78","added_by":"auto","created_at":"2025-09-30 08:17:48","extension":"jpeg","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":973682,"visible":true,"origin":"","legend":"\u003cp\u003eLithosphere structure training features (Irandoust et al., 2022) are \u003cstrong\u003ea\u003c/strong\u003e) averaged upper crustal shear velocity; \u003cstrong\u003eb\u003c/strong\u003e) averaged lower crustal shear velocity; \u003cstrong\u003ec\u003c/strong\u003e) averaged upper mantle shear velocity; and \u003cstrong\u003ed\u003c/strong\u003e) Moho depth.\u003c/p\u003e","description":"","filename":"image5.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7730637/v1/dc0d37bdd14fbc11d9f0a2b8.jpeg"},{"id":92484110,"identity":"c5e7e764-6c9c-4963-9683-30d904a19c66","added_by":"auto","created_at":"2025-09-30 08:17:48","extension":"jpeg","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":286170,"visible":true,"origin":"","legend":"\u003cp\u003eSimplified workflow of Mineral Prospectivity Mapping (MPM) approach. The process flows from left to right, beginning with the preparation and pre-processing of geospatial auxiliary feature layers (training features) and calculation of spatial density of known ore deposits (training target). Learning methods are evaluated to identify suitable algorithms (e.g., Linear Regression, SVM, Ensemble and Neural Networks). Feature importance is analyzed. Ensemble Bagged Trees with Neural Networks base models are selected due to their optimum performance. Hyperparameter tuning is applied to achieve optimize performance. Predictions from multiple base models are aggregated (e.g., via averaging or majority voting) to produce final mineral prospectivity maps highlighting zones with high mineral potential.\u003c/p\u003e","description":"","filename":"image6.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7730637/v1/b5ce91ef56ff2564475edf34.jpeg"},{"id":92486397,"identity":"0ac3a344-f161-4008-b130-82a0e1df9c0a","added_by":"auto","created_at":"2025-09-30 08:41:48","extension":"jpeg","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":343640,"visible":true,"origin":"","legend":"\u003cp\u003eThe importance of the training features determined by the Minimum Redundancy and Maximum Relevance (MRMR) algorithm (Ding and Peng, 2005).\u003c/p\u003e","description":"","filename":"image7.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7730637/v1/66cfd363b58177114db18598.jpeg"},{"id":92484989,"identity":"c9671111-3f74-42ae-8763-a3d7129470aa","added_by":"auto","created_at":"2025-09-30 08:25:48","extension":"jpeg","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":615303,"visible":true,"origin":"","legend":"\u003cp\u003eThe predicted spatial density of the ore deposits (i.e., metallogenic intensity) across the Iranian plateau, achieved by the Ensemble Bagged Trees method.\u003c/p\u003e","description":"","filename":"image8.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7730637/v1/bd600ab6bdde6d7d757f7ffd.jpeg"},{"id":92485228,"identity":"06a8d574-ec79-4197-b1f9-8eaf76dd4035","added_by":"auto","created_at":"2025-09-30 08:33:48","extension":"jpeg","order_by":9,"title":"Figure 9","display":"","copyAsset":false,"role":"figure","size":230802,"visible":true,"origin":"","legend":"\u003cp\u003eThe scatter plot compares the predicted spatial density of mineral deposits with the observation (Figure 2b). The prediction model was computed by the Ensemble Bagged Trees method, showing a high coefficient of correlation (R²) of 0.98.\u003c/p\u003e","description":"","filename":"image9.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7730637/v1/14add3f3aec665d6c361903b.jpeg"},{"id":92484994,"identity":"ebca8044-41fb-4782-87f1-8a75d359346a","added_by":"auto","created_at":"2025-09-30 08:25:48","extension":"jpeg","order_by":10,"title":"Figure 10","display":"","copyAsset":false,"role":"figure","size":658367,"visible":true,"origin":"","legend":"\u003cp\u003eThe residual density of the ore deposits (i.e., residual metallogenic intensity) derived by subtracting the predicted spatial density of the ore occurrence (Figure 8) from the observed spatial density of the ore occurrence (Figure 2b).\u003c/p\u003e","description":"","filename":"image10.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7730637/v1/29bbf13ca254691934451666.jpeg"},{"id":92486569,"identity":"b90a3ac2-86da-4335-8448-a6f0c1318dd2","added_by":"auto","created_at":"2025-09-30 08:49:51","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":7993728,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7730637/v1/7d9424bf-7d62-4422-beb6-3096adcdb770.pdf"},{"id":92484122,"identity":"ecdfc794-7ed5-4d19-b080-e579a95a70a4","added_by":"auto","created_at":"2025-09-30 08:17:48","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":6854466,"visible":true,"origin":"","legend":"","description":"","filename":"Supplementary.docx","url":"https://assets-eu.researchsquare.com/files/rs-7730637/v1/9fae352f230f2d38f9073a09.docx"}],"financialInterests":"The authors declare no competing interests.","formattedTitle":"\u003cp\u003e\u003cstrong\u003eMachine learning (ML)-based mineral prospectivity mapping (MPM): Detecting Iranian plateau high-potential metallogenic zones using geospatial big data\u003c/strong\u003e\u003c/p\u003e","fulltext":[{"header":"Highlights","content":"\u003cul\u003e\n \u003cli\u003eAn AI-aided approach was developed for mineral prospectivity mapping.\u003c/li\u003e\n \u003cli\u003eSixty-nine geological and geophysical variables were used for model training.\u003c/li\u003e\n \u003cli\u003eThe predicted spatial density of ore deposits was estimated across the Iranian plateau.\u003c/li\u003e\n \u003cli\u003eResidual spatial density anomalies highlight high-potential underexplored ore deposit zones.\u003c/li\u003e\n \u003cli\u003eThe study offers actionable insights for reducing future exploration risks and costs.\u003c/li\u003e\n\u003c/ul\u003e"},{"header":"1. Introduction","content":"\u003cp\u003eThe increasing global demand for new metallic ore resources is accompanied by declining discoveries in recent decades, since deposits with surface outcrops have mostly been discovered. It is speculated that unexplored deposits are either concealed beneath sediment cover or obscured by complex tectono-magmatic events (Davies et al., \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). Traditional exploration methods such as geological, geochemical, and geophysical surveying, and drilling are used to identify hidden resources (Zhao, \u003cspan citationid=\"CR135\" class=\"CitationRef\"\u003e2007\u003c/span\u003e; Dentith and Mudge, \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e2014\u003c/span\u003e; Adiri et al., \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Dentith et al., \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). However, the traditional methods are often expensive and time-consuming and face challenges in handling diverse big data formats with typically low resolution and inadequate survey coverage, particularly at the regional scale (Gonzalez-Alvarez et al., \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e2020\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eTo address these challenges and limitations, Mineral Prospectivity Mapping (MPM) has emerged as a fast, cost-effective, and data-driven approach for prioritizing potential metallogenic zones. The MPM approach leverages different geospatial data and insights from known deposits to predict new exploration targets on a regional scale. Formation of ore deposits is governed by various complex geological processes including magmatic differentiation, hydrothermal fluid circulation, sediment accumulation, metamorphic transformation, and supergene enrichment (Bierlein et al., \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2006\u003c/span\u003e; Groves and Bierlein, 2007; Kesler and Simon, \u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e2015\u003c/span\u003e). Despite substantial progress in understanding these processes, integration of diverse geospatial datasets into a comprehensive mineral exploration model remains challenging (McCuaig et al., \u003cspan citationid=\"CR75\" class=\"CitationRef\"\u003e2010\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eThe recently introduced artificial intelligence (AI) and machine learning (ML) algorithms have revolutionized geoscientific big data analysis (Bergen et al., \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; L\u0026ouml;sing and Ebbing, \u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Woodhead and Landry, \u003cspan citationid=\"CR122\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Guo and Yang, \u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Li et al., \u003cspan citationid=\"CR67\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Chukwu et al., \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Teknik, \u003cspan citationid=\"CR115\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Recent innovative AI advances are also revolutionizing mineral prospectivity modelling by enhancing processing ability, improving targeting accuracy, and optimizing exploration strategies while minimizing environmental impacts. AI-based methods have been successfully used for uncovering complex relationships between known mineral deposits and associated geological and geophysical datasets (Chen and Wu, \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Maepa et al., \u003cspan citationid=\"CR70\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Parsa and Carranza, \u003cspan citationid=\"CR89\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Parsa and Maghsoudi, \u003cspan citationid=\"CR90\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; N. Yang et al., \u003cspan citationid=\"CR126\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Yin et al., \u003cspan citationid=\"CR128\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Zuo and Xu, \u003cspan citationid=\"CR136\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Li et al., \u003cspan citationid=\"CR65\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Wake et al., \u003cspan citationid=\"CR121\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). ML-based MPM workflow employs statistical models to predict mineral potential by analyzing key attributes such as lithology, structural patterns, and geophysical anomalies. Then the trained ML models could be used to predict the likelihood of mineral deposits in less explored regions (Holden et al., \u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e2008\u003c/span\u003e; Carranza and Laborte, \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2015\u003c/span\u003e; Xiong and Zuo, \u003cspan citationid=\"CR124\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; McMillan et al., \u003cspan citationid=\"CR77\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Benaissi et al., \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). However, the challenges remain because of the scarcity of well-known deposits, their irregular distributions, as well as the complexity of ore formation, which complicate the development of universal models for different ore deposits (e.g., Xiong et al., \u003cspan citationid=\"CR125\" class=\"CitationRef\"\u003e2018\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eIn this study, we apply AI-based techniques to address the regional-scale exploration challenges on the Iranian plateau. Our approach investigates the statistical relationship between the spatial density of metallic deposits with geological and geophysical data. An integrated metallic mineral dataset was formed by compiling seven ore deposits of iron (Fe), copper (Cu), lead (Pb), zinc (Zn), gold (Au), silver (Ag), and magnesium (Mg). The spatial density of mineral deposits served as a target variable within a series of regression-based learning models, aiming to detect spatial patterns related to metallogenic potential. A novel transformation of complex vector-based geological maps into raster grid formats was introduced in this study to facilitate their integration into a geospatial big data framework suitable for machine learning analysis. We introduced an adaptive AI-aided workflow for regional-scale mineral prospectivity mapping that systematically evaluates the predictive performance of multiple regression algorithms to optimize the detection of the prospective zones across the Iranian plateau. The proposed methodology overcomes the heterogeneity and inconsistency of datasets and provides valuable guidance for prioritizing underexplored metallogenic zones. The approach provides a cost-effective solution with reduced financial risks, thereby facilitating more efficient and targeted mineral resource exploration efforts with minimum environmental impacts.\u003c/p\u003e"},{"header":"2. Geological framework of the Iranian plateau","content":"\u003cp\u003eThe Iranian plateau represents a complex and dynamic geological setting, where shaped by a series of temporally and spatially diverse geodynamic and tectonic events. Its evolution is closely linked to the sequential opening and closure of the Paleo- and Neo-Tethyan oceans, events that played a crucial role in the region's tectonostratigraphic architecture and metallogenic development (Stampfli and Borel, \u003cspan citationid=\"CR107\" class=\"CitationRef\"\u003e2002\u003c/span\u003e). Therefore, Iranian plateau is composed of continental blocks that are separated by Tethyan oceanic suture zones. The convergence between Arabian and Eurasian plates, accompanied by the progressive accretion of multiple continental microplates along the southern margin of Eurasia, where it has fundamentally shaped the present-day configuration of the Iranian plateau (Stocklin, \u003cspan citationid=\"CR109\" class=\"CitationRef\"\u003e1968\u003c/span\u003e; St\u0026ouml;cklin, \u003cspan citationid=\"CR108\" class=\"CitationRef\"\u003e1974\u003c/span\u003e; Berberian and King, \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e1981\u003c/span\u003e; Alavi, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e1996\u003c/span\u003e). The ore deposits' formation, their types and spatial distribution are closely associated with the magmatic history across the Iranian plateau, especially where the various phases of crustal extensions and compressions within the Tethyan orogeny, extending from the Early Palaeozoic to the Cenozoic (Stampfli, \u003cspan citationid=\"CR105\" class=\"CitationRef\"\u003e2000\u003c/span\u003e; Richards, \u003cspan citationid=\"CR97\" class=\"CitationRef\"\u003e2015\u003c/span\u003e). Many of the major crustal-scale lineaments were reactivated during Mesozoic and Cenozoic tectonic events and facilitated both extensive crustal extension and exhumation intrusions (Shahabpour, \u003cspan citationid=\"CR100\" class=\"CitationRef\"\u003e1999\u003c/span\u003e; Moritz et al., \u003cspan citationid=\"CR85\" class=\"CitationRef\"\u003e2006\u003c/span\u003e; Meshkani et al., \u003cspan citationid=\"CR79\" class=\"CitationRef\"\u003e2013\u003c/span\u003e; Bagheri, \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2015\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eThe exposures of Neoproterozoic-Early Cambrian crystalline basement rocks in the Iranian plateau and the trend of magmatic belts provide key controls on the distribution and localization of mineral deposits (Meshkani et al., \u003cspan citationid=\"CR79\" class=\"CitationRef\"\u003e2013\u003c/span\u003e). Extensional tectonics associated with the rifting of the Neo-Tethys ocean led to the detachment of the Cimmerian terranes from northern Gondwana during the Permian to Triassic. These Cimmerian terranes include the Anatolide-Tauride, Central Iran, Tibet, and Indochina. The northward drifting of these terranes ended with the event of their collision with the southern margin of Eurasia from the Late Triassic to the Early Jurassic, which led to the formation of the Paleo-Tethyan suture zone (Stampfli, \u003cspan citationid=\"CR105\" class=\"CitationRef\"\u003e2000\u003c/span\u003e; Stampfli and Borel, \u003cspan citationid=\"CR107\" class=\"CitationRef\"\u003e2002\u003c/span\u003e; Richards, \u003cspan citationid=\"CR97\" class=\"CitationRef\"\u003e2015\u003c/span\u003e). During the Mesozoic and Cenozoic, the subduction of the Neo-Tethys ocean beneath the Central Iran blocks formed fore-arc basins, extensive magmatism, and back-arc basins throughout the Iranian plateau (e.g., Monsef et al., \u003cspan citationid=\"CR84\" class=\"CitationRef\"\u003e2022\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eOne of the most significant magmatic arcs within the Iranian plateau is the Urumieh-Dokhtar magmatic arc (UDMA), which is marked by a subduction-related magmatism caused by the subduction of the Neo-Tethys ocean beneath the Central Iran block. The Urumieh-Dokhtar magmatic arc mostly formed by pronounced magmatic flare-ups throughout the Eocene to the Oligocene (ca. 55\u0026thinsp;\u0026minus;\u0026thinsp;25 Ma) (Verdel et al., \u003cspan citationid=\"CR120\" class=\"CitationRef\"\u003e2011\u003c/span\u003e; Chiu et al., \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2013\u003c/span\u003e). The magmatic flare-ups of UMDA were followed by widespread post-collisional extensional magmatism during the Neogene, which is attributed to the slab break-off or lithospheric delamination (e.g., Allen et al., \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2013\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eThe composite tectonic unit of the Central Iranian block consists of three major blocks of Lut, Tabas, and Yazd (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). Eastern Iran includes the Sistan suture zone and Lut block. Eastern Iran is characterized by intense Tertiary strike-slip faulting, widespread magmatism, and N-S trending ophiolitic belts (e.g., Omidianfar et al., \u003cspan citationid=\"CR88\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). Subduction-related magmatism associated with these tectonic processes has contributed significantly to the metallogenic evolution of Eastern Iran (Arjmandzadeh et al., \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2011\u003c/span\u003e; Alaminia et al., \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2013\u003c/span\u003e; Richards, \u003cspan citationid=\"CR97\" class=\"CitationRef\"\u003e2015\u003c/span\u003e). In the northeast of Central Iran and north of Eastern Iran, the E-W trending Sabzevar magmatic-ophiolitic belt extends about 700 km along the northern side of the Dorouneh fault. This belt is bound to the Kopeh-Dagh zone (Eurasian plate) in its north. The Kopeh-Dagh Mountains mark the most northeastern edge of the deformation zone of the Arabia-Eurasia collisional zone in the Iranian plateau. The Kopeh-Dagh is marked with a\u0026thinsp;~\u0026thinsp;10 km deep folded Mesozoic-Tertiary sediment sequence. The Sistan and Sabzevar ophiolites represent the evidence of Neo-Tethys back-arc basins and serve as an important lithological marker delineating the boundaries of major tectonic units (St\u0026ouml;cklin, \u003cspan citationid=\"CR110\" class=\"CitationRef\"\u003e1968\u003c/span\u003e; Richards, \u003cspan citationid=\"CR97\" class=\"CitationRef\"\u003e2015\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eDuring the Neogene, the Arabian plate collided with the Central Iran block, leading to the formation of the Bitlis-Zagros suture zone (St\u0026ouml;cklin, \u003cspan citationid=\"CR110\" class=\"CitationRef\"\u003e1968\u003c/span\u003e; Berberian and King, \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e1981\u003c/span\u003e; Boulin, \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e1991\u003c/span\u003e; Stampfli and Borel, \u003cspan citationid=\"CR107\" class=\"CitationRef\"\u003e2002\u003c/span\u003e). The continuous northward convergence of the Arabian plate gave rise to the Zagros fold and thrust belt with an extensive NW-SE-trending orogenic belt in the south of the Zagros suture zone. The Zagros orogenic system remains tectonically active, particularly where the Main Recent Fault (MRF) and the Main Zagros Thrust (MZT) consume part of the regions' ongoing continental convergence (Stampfli and Borel, \u003cspan citationid=\"CR106\" class=\"CitationRef\"\u003e2004\u003c/span\u003e; Richards, \u003cspan citationid=\"CR97\" class=\"CitationRef\"\u003e2015\u003c/span\u003e; Berberian, 1995; Sepehr and Cosgrove, 2004) (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). The Zagros suture zone lies approximately along MRF and MZT. The Zagros ophiolites are discontinuously exposed along the Zagros suture zone. The ophiolites can be divided into two parallel belts of the inner and outer Zagros ophiolitic belts. In the southwest periphery of the Central Iran block, the inner Zagros ophiolitic belt comprises Nain, Dehshir, and Baft ophiolites. The outer Zagros ophiolitic belt, along the Main Zagros Thrust, comprises Kurdistan, Kermanshah, Neyriz, and Hajiabad ophiolites (e.g., Monsef et al., \u003cspan citationid=\"CR82\" class=\"CitationRef\"\u003e2018\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eThe southeastern continuation of the Zagros fold and thrust belt identifies the Makran accretionary prism that extends for ~\u0026thinsp;900 km from southeastern Iran to southeastern Pakistan. In this region, the last piece of Arabian oceanic lithosphere is currently being subducted beneath the Makran accretionary complex in the south of Iran and Pakistan. The Makran subduction zone, which exhibits an anomalously low level of magmatism, is associated with the development of the Cenozoic Jazmurian Basin, interpreted as an old back-arc basin (Glennie et al., \u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e1990\u003c/span\u003e; Shahabpour, 2010; Penney et al., \u003cspan citationid=\"CR93\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Burg, \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Teknik et al., \u003cspan citationid=\"CR116\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). The Jazmurian basin is covered with a thick sedimentary cover reaching its maximum thickness of ~\u0026thinsp;20 km in its eastern edge (Enayat and Ghods, \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Mehrdar et al., \u003cspan citationid=\"CR78\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). The Zagros ophiolites continue southeast, toward the Makran ophiolites. The Makran ophiolites can also be divided into two distinct belts, consisting of the inner and outer Makran ophiolitic belts (e.g., Monsef et al., \u003cspan citationid=\"CR83\" class=\"CitationRef\"\u003e2019\u003c/span\u003e).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e"},{"header":"3. Ore deposits of the Iranian plateau","content":"\u003cp\u003eDifferent world-class ore deposits occurred in the Iranian plateau, making the Iranian plateau one of the most important metallogenic zones in western Asia and central Tethys (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e and Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). The most prominent ore deposits on the Iranian plateau are iron (Fe), copper (Cu), lead-zinc (Pb-Zn), and gold (Au) deposits. This is indebted to the complex geologic setting of the plateau, manifested by a series of tectonic, magmatic, hydrothermal, and metamorphic events along the major active geological boundaries. These processes have facilitated the formation of distinct metallogenic zones. Each zone is particularly associated with the distinct tectonic structures such as faults, shear zones, and suture boundaries, as well as geodynamic events such as subduction, continental collision, and post-collisional extension. The investigation of ore deposits was done using various methods such as remote sensing, geochemical analyses, and structural surveys (e.g., Zarasvandi et al., \u003cspan citationid=\"CR130\" class=\"CitationRef\"\u003e2005\u003c/span\u003e; Golmohammadi et al., \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e2015\u003c/span\u003e; TaleFazel et al., \u003cspan citationid=\"CR114\" class=\"CitationRef\"\u003e2019\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eThe influence of the geodynamic processes on the genesis of the Iranian ore deposits is addressed by comprehensive geological studies. In this context, the evolution of Tethyan oceans and associated tectono-magmatic activities caused the distribution of ore deposits within the Iranian plateau. Indeed, the sequential divergent and convergent events of the Tethyan oceans provide a suitable geological framework for the formation of a variety of ore deposits (e.g., porphyry, skarn, epithermal, MVT, VMS, and SEDEX). Igneous rocks, including volcanic, sub-volcanic, and plutonic bodies, crystallized from the magmas that derived from the partial melting of the upper mantle and then differentiated in crustal magma chambers. The metalliferous hydrothermal fluids exsolved from the evolved melts caused Fe, Cu, Pb-Zn, and Au mineralisation in the form of massive ores, breccias, veins, and veinlets at the shallow levels.\u003c/p\u003e\u003cp\u003eThe subduction of Tethyan oceans beneath the Iranian continental fragments produced arc-related and extensional back-arc basin magmatism with mostly calc-alkaline to alkaline affinities. After the collision between the Arabia and Central Iran or Lut and Afghan blocks, the lithospheric delamination and subsequent asthenospheric upwelling led to decompression melting of previously metasomatized sub-continental lithospheric mantle and the generation of high-K alkaline to shoshonitic volcanism and plutonic bodies. Porphyry deposits in Iran are intrusion-related deposits that are products of magmatic-hydrothermal activity at shallow crustal levels in subduction and post-collisional tectonic settings. Primarily copper, but also molybdenum and gold occur closely related to epizonal intrusions of porphyric magmatic rocks (e.g., Zarasvandi et al., \u003cspan citationid=\"CR130\" class=\"CitationRef\"\u003e2005\u003c/span\u003e, \u003cspan citationid=\"CR133\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Hezarkhani, \u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e2006\u003c/span\u003e; Aghazadeh et al., \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2015\u003c/span\u003e). Skarn-type deposits in Iran are formed by metasomatic replacement of carbonate rocks by hydrothermal fluids derived from granitoid bodies in subduction and collisional tectonic settings. They are characterized by the presence of copper and iron (e.g., Golmohammadi et al., \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e2015\u003c/span\u003e; Hassanpour and Rajabpour, \u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). Epithermal deposits in Iran, ranging from high-sulfidation to low-sulfidation types, are connected with terrestrial volcanism that formed by near-surface magmatic-hydrothermal processes circulating through fault systems and veins. They are often hosted within volcanic and volcaniclastic rocks remarkable as a major source of gold, silver, copper, lead, and zinc (e.g., Mehrabi et al., \u003cspan citationid=\"CR80\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; TaleFazel et al., \u003cspan citationid=\"CR114\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). Mississippi Valley-type (MVT) deposits in Iran are a type of epigenetic carbonate-hosted sulfide ore deposit, primarily known for their lead and zinc mineralization. These deposits are characterized by their migration of ore-forming fluids within fault and karst systems in a collisional tectonic setting (e.g., Rajabi et al., \u003cspan citationid=\"CR96\" class=\"CitationRef\"\u003e2013\u003c/span\u003e; Qaderi et al., \u003cspan citationid=\"CR95\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Volcanogenic massive sulphide (VMS) deposits in Iran are formed by submarine volcanic activity, or more precisely, seafloor hydrothermal activity related to the arc/intra-arc rifts and back-arc basins. The most important metals in VMS are copper, lead, and zinc, with trace contents of gold and silver that are hosted in volcano-sedimentary succession (e.g., Hajsadeghi et al., \u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Mousivand et al., \u003cspan citationid=\"CR86\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). Sedimentary exhalative (SEDEX) deposits in Iran are very similar in genesis to the VMS deposits that are primarily known for containing often lead and zinc with subordinate silver and iron formed by the precipitation of metal sulfides from hydrothermal fluids onto the seafloor sediments of arc/intra-arc rifts and back-arc basins (e.g., Maghfouri and Hosseinzadeh, \u003cspan citationid=\"CR71\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Maghfouri et al., \u003cspan citationid=\"CR72\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). Accordingly, these deposits within the Iranian plateau are formed in an integral ore-forming system and have a genetic link with magmatic-hydrothermal processes. These magmatic-hydrothermal activities have a relationship to the evolution of the Tethyan oceans.\u003c/p\u003e\u003cp\u003eCentral Iran, particularly the Bafgh region, is recognized as one of the most metallogenically significant zones in the Tethys belt, especially for its exceptional concentration of iron ore deposits. The region hosts some of the largest iron ore bodies within the Iranian plateau, such as the Choghart and Sechahoon deposits. These gigantic deposits are primarily associated with Neoproterozoic metamorphic and igneous complexes (Ghorbani, \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e2013a\u003c/span\u003e; Korehie et al., \u003cspan citationid=\"CR63\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). High aeromagnetic anomalies around the Bafgh region imply a high potential for further iron metallogenic zones (e.g., Torab and Lehmann, \u003cspan citationid=\"CR119\" class=\"CitationRef\"\u003e2007\u003c/span\u003e). Iron ore spatial distribution extends beyond Central Iran into the Sanandaj-Sirjan zone, Eastern Iran, and the Urumieh-Dokhtar magmatic arc. It includes a variety of deposit types such as magmatic, skarn, volcanogenic, and sedimentary types (e.g., Stosch et al., \u003cspan citationid=\"CR111\" class=\"CitationRef\"\u003e2011\u003c/span\u003e; Nabatian et al., \u003cspan citationid=\"CR87\" class=\"CitationRef\"\u003e2015\u003c/span\u003e; Hassanlouei and Rajabzadeh, \u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e2019\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eNumerous porphyry and skarn-type cupper ore deposits are predominantly associated with the Urumieh-Dokhtar magmatic arc. Significant clusters of porphyry copper deposits are found in the Jiroft-Shahrebabak zone (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ed), where the world-known porphyry copper deposits of Sarcheshmeh and Meiduk have been discovered (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ea). A similar tectonomagmatic activity and emplacement of magmatic intrusions along Urumieh-Dokhtar magmatic arc formed the Mazraeh and Sungon copper deposits (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ea), together with a cluster of minor deposits along the KaraDagh zone in northwestern Iran (Zarasvandi et al., \u003cspan citationid=\"CR131\" class=\"CitationRef\"\u003e2007\u003c/span\u003e; Hezarkhani, \u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e2008\u003c/span\u003e; Mollai et al., \u003cspan citationid=\"CR81\" class=\"CitationRef\"\u003e2009\u003c/span\u003e; Asadi et al., \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2014\u003c/span\u003e; Kheyrollahi et al., \u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e2018\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eLead-zinc deposits are widespread across the Sanandaj-Sirjan zone and more specifically the Malayer-Isfahan zone. Other major lead-zinc metallogenic zones are situated in Central Iran and the Alborz regions (Ghorbani, \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e2013b\u003c/span\u003e). Considering the diverse tectono-magmatic and metamorphic events, the various metallic ore deposits have been categorized based on their spatial distributions and types (Ghorbani, \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e2013b\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eThe gold deposits on the Iranian plateau are relatively less explored. However, the distribution of gold deposits is primarily extended within the Sanandaj-Sirjan metamorphic zone as well as along the Urumieh-Dokhtar magmatic arc, Azerbaijan-Alborz, and Eastern Iranian magmatic belts. These regions host epithermal, porphyry-related, and orogenic deposit types, reflecting the complex geodynamic evolution of the Tethyan metallogenic belt across the Iranian plateau (e.g., Moritz et al., \u003cspan citationid=\"CR85\" class=\"CitationRef\"\u003e2006\u003c/span\u003e; Richards et al., \u003cspan citationid=\"CR98\" class=\"CitationRef\"\u003e2006\u003c/span\u003e; Daliran, \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2008\u003c/span\u003e; Geranian et al., \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). The silver deposits, similar to gold deposits, are typically connected to the copper metallogenic zones. The spatial distribution of the magnesium mostly follows the trends of the Sistan suture zones (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ea).\u003c/p\u003e\u003cp\u003eIn this study, the metallogenic zones delineated by Ghorbani (\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e2013a\u003c/span\u003e) have been refined and updated by calculating the spatial density of mineral deposits (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ed). Regions exhibiting high spatial density (\u0026gt;\u0026thinsp;0.2 num/sq km) of mineral deposits have been identified. Consequently, the major metallogenic zones on the Iranian plateau are: (1) KaraDagh; (2) Takab-Tarom-Hashtjin; (3) Malayer-Isfahan; (4) Kashan-Natanz; (5) Toroud; (6) Anarak; (7) QaleBala; (8) Bafgh; (9) Khaf; (10) Nehbandan-Ferdous; (11) Jiroft-Shahrebabak; and (12) Kahnuj-Fanuj.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eList of major known Cu, Fe, Au, and Pb-Zn ore deposits throughout the Iranian plateau (after Meshkani et al., \u003cspan citationid=\"CR79\" class=\"CitationRef\"\u003e2013\u003c/span\u003e and references therein). The selected deposits here are grouped for qualitative evaluation of the AI-based models. Abbreviations: MVT, Mississippi Valley type deposit; VMS, Volcanogenic massive sulphide deposit; SEDEX, Sedimentary exhalative deposit.\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"10\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c9\" colnum=\"9\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c10\" colnum=\"10\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003eDeposits\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eLong\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003eLat\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003eHost/country rocks\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c5\"\u003e\u003cp\u003eStratigraphic age\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c6\"\u003e\u003cp\u003eMajor Commodity\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c7\"\u003e\u003cp\u003eOther accessory minerals\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c8\"\u003e\u003cp\u003eGenetic type\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colspan=\"2\" nameend=\"c10\" namest=\"c9\"\u003e\u003cp\u003eTonnage and grade\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMasjed daghi\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e46.936\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e38.875\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eGranite and andesitic rocks\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eOligo-Miocene\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eCu\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eAu\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003ePorphyry\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c10\" namest=\"c9\"\u003e\u003cp\u003e10 Mt \u0026minus;\u0026thinsp;0.7% Cu, 0.5 g/t Au\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMazraeh\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e46.983\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e38.625\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eVolcano-sedimentary rocks, granite, and limestone\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eOligo-Miocene\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eCu\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003ePb, Zn, Au, Ag\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003eSkarn\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c10\" namest=\"c9\"\u003e\u003cp\u003e1 Mt \u0026minus;\u0026thinsp;1.7% Cu, 0.3 g/t Au\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSungon\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e46.717\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e38.700\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eMonzodiorite and volcano-sedimentary rocks\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eOligo-Miocene\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eCu\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eMo, Pb, Zn, Au, Ag\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003ePorphyritic-skarn\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c10\" namest=\"c9\"\u003e\u003cp\u003e290 Mt \u0026minus;\u0026thinsp;0.76% Cu, 0.015% Mo\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eEnjerd\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e46.933\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e38.683\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eAndesite, granite, and limestone\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eCretaceous, Oligo-Miocene\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eCu\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eAu, Fe\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003eSkarn\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c10\" namest=\"c9\"\u003e\u003cp\u003e300Mt \u0026minus;\u0026thinsp;0.85% Cu\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAstamal\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e46.375\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e38.567\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eAndesite and tuff\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eOligo-Miocene\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eCu\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eMo, Pb, Zn\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003eVein\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c10\" namest=\"c9\"\u003e\u003cp\u003e\u0026gt;\u0026thinsp;1 Mt\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eJarou\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e50.550\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e35.708\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eAndesite and andesitic basalt\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eEocene-Oligocene\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eCu\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003ePb, Zn\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003eVein\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c10\" namest=\"c9\"\u003e\u003cp\u003e\u0026gt;\u0026thinsp;1 Mt\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eVeshnaveh\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e50.983\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e34.233\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eAndesitic basalt and tuff\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eEocene\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eCu\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003eVein\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c10\" namest=\"c9\"\u003e\u003cp\u003e\u0026gt;\u0026thinsp;1 Mt\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eDeh madan\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e51.083\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e31.600\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eLimestone, dolomite, and sandstone\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eCambrian\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eCu\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eZn, Pb, Co\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003eMVT\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c10\" namest=\"c9\"\u003e\u003cp\u003e\u0026gt;\u0026thinsp;1 Mt\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMeskani\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e53.450\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e33.325\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eTrachyandesite and basalt\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eEocene\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eCu\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eNi, Co, U, Bi, Au, Pb, Zn, Ag\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003eHydrothermal (volcanogenic)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c10\" namest=\"c9\"\u003e\u003cp\u003e~\u0026thinsp;1 Mt.- 2%Cu, 002% Ni, 15 g/t Ag\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eTalmesi\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e53.450\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e33.383\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eTrachyandesite and basalt\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eEocene\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eCu\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eNi, Co, U, Bi, Au, Pb, Zn, Ag\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003eHydrothermal (volcanogenic)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c10\" namest=\"c9\"\u003e\u003cp\u003e~\u0026thinsp;1 Mt \u0026minus;\u0026thinsp;2.2% Cu, 002% Ni\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAli abad\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e48.983\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e36.508\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003ePorphyritic granodiorite and tuff\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eOligo-Miocene\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eCu\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eMo\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003ePorphyry\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c10\" namest=\"c9\"\u003e\u003cp\u003e40 Mt \u0026minus;\u0026thinsp;0.7% Cu, 0.005% Mo\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eDarreh zereshk\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e53.842\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e31.567\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eGranodiorite, andesitic tuff and limestone\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eEocene-Miocene\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eCu\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003ePorphyry\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c10\" namest=\"c9\"\u003e\u003cp\u003e23 Mt \u0026minus;\u0026thinsp;0.68% Cu, 0.01% Mo\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eChah mousa\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e54.867\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e35.475\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eAndesitic tuff and lava\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003ePaleogene\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eCu\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003eVein\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c10\" namest=\"c9\"\u003e\u003cp\u003e\u0026gt;\u0026thinsp;1 Mt\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eTaknar\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e57.783\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e35.367\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eRhyolite and schist\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eLate Paleozoic\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eCu\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eAg, Au, Pb, Zn\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003eSEDEX\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c10\" namest=\"c9\"\u003e\u003cp\u003e\u0026gt;\u0026thinsp;2 Mt \u0026minus;\u0026thinsp;3% Cu, 1.5% Zn, 1.%Pb\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eQaleh zari\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e58.955\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e32.362\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eAndesitic to basaltic lava and tuff\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eEocene\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eCu\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eFe, Au, Ag, Pb, Zn\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003eHydrothermal\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c10\" namest=\"c9\"\u003e\u003cp\u003e1.3 Mt \u0026minus;\u0026thinsp;3% Cu, 2 g/t Au, 30 g/t Ag\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMeiduk\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e55.200\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e30.467\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eAndesitic basalt and pyroclastic rocks\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eMiocene\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eCu\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003ePorphyry\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c10\" namest=\"c9\"\u003e\u003cp\u003e500 Mt \u0026minus;\u0026thinsp;0.83% Cu, 0.01% Mo\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eDarreh-zerreshk\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e55.917\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e29.883\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eVolcano-sedimentary rocks and porphyritic diorite\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eMiocene\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eCu\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eMo, Pb, Zn\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003ePorphyry\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c10\" namest=\"c9\"\u003e\u003cp\u003e\u0026gt;\u0026thinsp;1 Mt 0.64%, 0.004% Mo\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSarcheshmeh\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e55.867\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e29.950\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003ePorphyritic granodiorite and andesite\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eEocene-Miocene\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eCu\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eMo, Au\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003ePorphyry\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c10\" namest=\"c9\"\u003e\u003cp\u003e1200 Mt \u0026minus;\u0026thinsp;0.7% Cu, 0.03% Mo, 0.08 g/t Au, 3 g/t Ag\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eChahar gonbad\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e56.183\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e29.592\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003ePorphyritic quartz diorite and andesitic tuff\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eEocene-Oligo-Miocene\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eCu\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003ePb, Zn, Au, Ag\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003eHydrothermal\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c10\" namest=\"c9\"\u003e\u003cp\u003e3 Mt \u0026minus;\u0026thinsp;1.67% Cu\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSheikh aali\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e56.758\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e28.133\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003ePillow lava and volcanic rocks\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eUpper Cretaceous\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eCu\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eAu\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003eVMS\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c10\" namest=\"c9\"\u003e\u003cp\u003e\u0026gt;\u0026thinsp;1 Mt \u0026minus;\u0026thinsp;2% Cu, 64 g/t Au\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eRameshk\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e58.817\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e26.817\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eGabbro, andesitic basalt and limestone\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eUpper Cretaceous-Lower Paleocene\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eCu\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003eVMS?\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c10\" namest=\"c9\"\u003e\u003cp\u003e\u0026gt;\u0026thinsp;1 Mt\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAnguran\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e47.406\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e36.628\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eLimestone, micaschist, and marble\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eProterozoic\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003ePb, Zn\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eAg\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c9\" namest=\"c8\"\u003e\u003cp\u003eCarbonate-hosted\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e22 Mt \u0026minus;\u0026thinsp;24% Zn, 6% Pb\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAlam kandi\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e47.283\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e36.717\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eSchist, marble, quartzite, and tuff\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eProterozoic\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003ePb, Zn\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eCu\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c9\" namest=\"c8\"\u003e\u003cp\u003eCarbonate-hosted\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e~\u0026thinsp;1 Mt \u0026minus;\u0026thinsp;7% Zn, 3% Pb\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAhangaran\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e48.991\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e34.186\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eSandy dolomite, quartzite and shale\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eLower Cretaceous\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003ePb, Zn\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eFe, Ba, Cd, Ag\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c9\" namest=\"c8\"\u003e\u003cp\u003eMVT\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e2Mt \u0026minus;\u0026thinsp;3% Pb, 1% Zn, 200 g/t Ag\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eEmarat\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e49.603\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e33.856\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eLimestone and dolomite\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eCretaceous\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003ePb, Zn\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eCu, Ag, Cd\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c9\" namest=\"c8\"\u003e\u003cp\u003eMVT\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e10 Mt \u0026minus;\u0026thinsp;2% Pb, 3% Zn\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eDona\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e51.450\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e36.165\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eLimestone and dolomite\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003ePermian\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003ePb, Zn\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eBa, Ag\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c9\" namest=\"c8\"\u003e\u003cp\u003eCarbonate-hosted\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e6.5 Mt \u0026minus;\u0026thinsp;5% Pb, 1% Zn, 150 g/t Ag\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eDarreh noqreh\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e50.217\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e33.525\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eLimestone, pyroclastic and volcanic rocks\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eCretaceous\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003ePb, Zn\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eCu, Ag\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c9\" namest=\"c8\"\u003e\u003cp\u003eMVT\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e~\u0026thinsp;1 Mt \u0026minus;\u0026thinsp;21% Pb, 2% Zn, 150 g/t Ag\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eLakan\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e50.648\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e33.109\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eSilicified limestone and shale\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eCretaceous\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003ePb, Zn\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eCu, Ag, Ba\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c9\" namest=\"c8\"\u003e\u003cp\u003eSEDEX\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e5 Mt \u0026minus;\u0026thinsp;3% Zn, 4.5% Pb\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eHosein abad\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e50.983\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e32.958\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eBlack shale and sandstone\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eJurassic\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003ePb, Zn\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c9\" namest=\"c8\"\u003e\u003cp\u003eSEDEX\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e2 Mt \u0026minus;\u0026thinsp;1% Zn, 4% Pb\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAnjeereh tiran\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e51.125\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e32.745\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eDolomite, limestone, and shale\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eCretaceous\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003ePb, Zn\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eCu, Ag, Cd, Sb\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c9\" namest=\"c8\"\u003e\u003cp\u003eMVT\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e1.5 Mt \u0026minus;\u0026thinsp;4% Zn, 1% Pb\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eIrankuh\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e51.625\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e32.500\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eDolomite, limestone, and shale\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eCretaceous\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003ePb, Zn\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eAg, Cd\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c9\" namest=\"c8\"\u003e\u003cp\u003eMVT\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e17 Mt \u0026minus;\u0026thinsp;11% Zn, 2.5% Pb\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eKuh-e-surmeh\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e52.517\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e28.500\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eDolomite, limestone, and sandstone\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eLower Paleozoic\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003ePb, Zn\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c9\" namest=\"c8\"\u003e\u003cp\u003eMVT\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e~\u0026thinsp;1 Mt \u0026minus;\u0026thinsp;17% Zn, 2%Pb\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eNakhlak\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e53.839\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e33.564\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eLimestone, shale, sandstone, and marl\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eMiddle Triassic-Upper Cretaceous\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003ePb, Zn\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eAg\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c9\" namest=\"c8\"\u003e\u003cp\u003eMVT\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e3 Mt \u0026minus;\u0026thinsp;5% Zn, 75 g/t Ag\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMoujen\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e54.658\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e36.525\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eLimestone\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003ePermo-Triassic\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003ePb, Zn\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eFe\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c9\" namest=\"c8\"\u003e\u003cp\u003eMVT\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e~\u0026thinsp;1 Mt \u0026minus;\u0026thinsp;2% Pb, 4% Zn\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eKhan jar\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e54.558\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e35.367\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eDolomite and limestone\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eCretaceous\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003ePb, Zn\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eAg\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c9\" namest=\"c8\"\u003e\u003cp\u003eCarbonate-hosted\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e1 Mt \u0026minus;\u0026thinsp;20% Pb, 4% Zn, 300 g/t Ag\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eChah sorb\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e56.617\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e34.050\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eDolomite and limestone\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eMiddle Triassic\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003ePb, Zn\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eAg\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c9\" namest=\"c8\"\u003e\u003cp\u003eCarbonate-hosted\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e~\u0026thinsp;1 Mt \u0026minus;\u0026thinsp;5% Zn, 2.5%Pb\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eOzbak kuh\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e57.117\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e34.667\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eLimestone, shale, and sandstone\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003ePaleozoic\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003ePb, Zn\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c9\" namest=\"c8\"\u003e\u003cp\u003eCarbonate-hosted\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e2 Mt \u0026minus;\u0026thinsp;10% Zn, 4%Pb\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMehdi abad\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e55.025\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e31.483\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eLimestone, dolomite, and schist\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eCretaceous\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003ePb, Zn\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eFe, Ba\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c9\" namest=\"c8\"\u003e\u003cp\u003eMVT\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e218 Mt \u0026minus;\u0026thinsp;7% Zn, 2.3% Pb, 51 g/t Ag\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eKoushk\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e55.775\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e31.733\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eBlack shale\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eProterozoic\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003ePb, Zn\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eAg\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c9\" namest=\"c8\"\u003e\u003cp\u003eSEDEX\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e5 Mt \u0026minus;\u0026thinsp;15% Zn, 3%Pb\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eChah mir\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e56.042\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e31.65\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eBlack siltstone\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eLate Cambrian\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003ePb, Zn\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c9\" namest=\"c8\"\u003e\u003cp\u003eSEDEX\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e~\u0026thinsp;1 Mt \u0026minus;\u0026thinsp;6% Zn, 3.5%Pb\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAghdarreh\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e47.017\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e36.667\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eSilicified limestone\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eLower Miocene\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eAu\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eSb, As, Hg\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c9\" namest=\"c8\"\u003e\u003cp\u003eSediment hosted\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e\u0026gt;\u0026thinsp;5 Mt \u0026minus;\u0026thinsp;4.5 g/t Au\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eZarshuran\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e47.133\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e36.725\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eDolomitic limestone and black shale\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eUpper Proterozoic-Lower Cambrian\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eAu\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eAs, Sb, Hg, Zn\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c9\" namest=\"c8\"\u003e\u003cp\u003eSediment hosted\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e12 Mt \u0026minus;\u0026thinsp;7.9 g/t Au\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eTouzlar\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e47.466\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e36.833\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eVolcanic rocks\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eOligo-Miocene\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eAu\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c9\" namest=\"c8\"\u003e\u003cp\u003eLow sulfidation epithermal\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e~\u0026thinsp;1 Mt \u0026minus;\u0026thinsp;3.1 g/t Au\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eKervian\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e46.100\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e36.133\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eMeta-volcanic and schist\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eMesozoic\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eAu\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c9\" namest=\"c8\"\u003e\u003cp\u003eOrogenic gold\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e~\u0026thinsp;1 Mt \u0026minus;\u0026thinsp;3 g/t Au\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAlut\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e45.597\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e36.147\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eQuartz schist\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eMesozoic\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eAu\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c9\" namest=\"c8\"\u003e\u003cp\u003eOrogenic gold\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e~\u0026thinsp;1 Mt \u0026minus;\u0026thinsp;2.5 g/t Au\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSari gunay\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e48.092\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e35.183\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003ePorphyritic micro-diorite and rhyolite\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eTertiary\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eAu\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eSb, Cu, As, Hg\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c9\" namest=\"c8\"\u003e\u003cp\u003eHigh sulfidation epithermal\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e108 Mt \u0026minus;\u0026thinsp;2.3 g/t Au\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAstaneh\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e49.325\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e33.867\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eMicro-granite\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eTriassic-Jurassic\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eAu\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eCu, W\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c9\" namest=\"c8\"\u003e\u003cp\u003eIntrusion-related gold\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e~\u0026thinsp;1 g/t Au\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMuteh\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e50.608\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e33.667\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eSchist and meta-rhyolite\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eUpper Proterozoic-Lower Cambrian\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eAu\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eCu\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c9\" namest=\"c8\"\u003e\u003cp\u003eEpithermal\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e8 Mt \u0026minus;\u0026thinsp;3.2 g/t Au\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eZarrin\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e54.625\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e32.675\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eAlluvium\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eQuaternary\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eAu\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eW\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c9\" namest=\"c8\"\u003e\u003cp\u003ePlacer\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e1 Mt - \u0026lt; 1 g/t Au\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eKuh-e-zar\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e54.650\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e35.467\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eAlluvium\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eQuaternary\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eAu\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c9\" namest=\"c8\"\u003e\u003cp\u003ePlacer\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e~\u0026thinsp;1 Mt \u0026minus;\u0026thinsp;.05 g/t Au\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eGandi\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e54.633\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e35.317\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eVolcano-sedimentary rocks\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eEocene\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eAu\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eCu, Pb, Zn, Ag\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c9\" namest=\"c8\"\u003e\u003cp\u003eIntermediate epithermal\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e~\u0026thinsp;1 Mt \u0026minus;\u0026thinsp;5 g/t Au\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eZar mehr\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e58.927\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e35.174\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eAndesite and granodiorite\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eEocene\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eAu\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c9\" namest=\"c8\"\u003e\u003cp\u003eIOGC\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e1.5 Mt \u0026minus;\u0026thinsp;4 g/t Au\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eZartorosht\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e57.211\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e28.219\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eGreenschist\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003ePaleozoic\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eAu\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c9\" namest=\"c8\"\u003e\u003cp\u003eOrogenic gold type\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e\u0026gt;\u0026thinsp;1 Mt \u0026minus;\u0026thinsp;3 g/t Au\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eShahrak\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e47.828\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e36.420\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eLimestone, rhyodacite, and andesite\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eOligo-Miocene\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eFe\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003eVolcanogenic\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c10\" namest=\"c9\"\u003e\u003cp\u003e100 Mt \u0026minus;\u0026thinsp;57% Fe\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eShams abad\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e49.725\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e33.817\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eSandy dolomite and shale\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eCretaceous\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eFe\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eMn, Cu\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003eVolcanogenic\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c10\" namest=\"c9\"\u003e\u003cp\u003e30 Mt \u0026minus;\u0026thinsp;47% Fe, 4% Mn\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSangan\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e60.400\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e34.408\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eGranodiorite, limestone and schist\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eProterozoic\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eFe\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003eSkarn\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c10\" namest=\"c9\"\u003e\u003cp\u003e900 Mt \u0026minus;\u0026thinsp;47% Fe\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eRobat posht badam\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e55.567\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e32.958\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eGneiss, amphibolite, and marble\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eTriassic\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eFe\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003eMagmatic\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c10\" namest=\"c9\"\u003e\u003cp\u003e\u0026gt;\u0026thinsp;1 Mt\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eChador Malu\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e55.500\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e32.300\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eMeta-syenite, schist, and rhyolite\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eProterozoic\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eFe\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003eMagmatic\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c10\" namest=\"c9\"\u003e\u003cp\u003e450 Mt \u0026minus;\u0026thinsp;56% Fe\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSechahoon\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e55.643\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e31.718\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eMetasomatic granite, andesite, and tuff\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eProterozoic\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eFe\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003eMagmatic\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c10\" namest=\"c9\"\u003e\u003cp\u003e132 Mt \u0026minus;\u0026thinsp;35% Fe\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eChoghart\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e55.467\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e31.700\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eSyenite, schist, and rhyolite\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eProterozoic\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eFe\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eMn\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003eMagmatic\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c10\" namest=\"c9\"\u003e\u003cp\u003e350 Mt \u0026minus;\u0026thinsp;55% Fe\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eGol Gohar\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e55.083\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e29.267\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eSchist, marble, and quartzite\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eProterozoic\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eFe\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003eSkarn\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c10\" namest=\"c9\"\u003e\u003cp\u003e1200 Mt \u0026minus;\u0026thinsp;55% Fe\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eTang-E-Zagh\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e56.017\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e27.950\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eLimestone and marl\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eProterozoic\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eFe\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eMn\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003eSedimentary\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c10\" namest=\"c9\"\u003e\u003cp\u003e5 Mt \u0026minus;\u0026thinsp;42% Fe, 1% Mn\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e"},{"header":"4. Training dataset","content":"\u003cp\u003eThe construction of the training database is conducted through the following multi-stage workflow: \u003cb\u003e1)\u003c/b\u003e Compilation of publicly available geological and geophysical datasets, which are collected from authoritative references (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e); \u003cb\u003e2)\u003c/b\u003e Feature enhancement through the use of edge-detection and gradient-based filters, which provide additional training features. These features highlight geological contrast and structural delineation; \u003cb\u003e3)\u003c/b\u003e Target definition is a parameter that represents a metallogenic proxy. Here, the training target was generated based on the spatial density of mineral occurrence. This parameter illustrates the intensity and distribution of metallogenic intensity (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e); \u003cb\u003e4)\u003c/b\u003e The training dataset was formed by compiling all training input databases (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e), which were sampled ore deposit locations (collected data in step 1). The training dataset represents the relevance of the geological or geophysical features to the training target.\u003c/p\u003e\u003cp\u003eThe original compiled training dataset consists of 2351 sample points. Before any process, a certain data quality checking and data cleaning are required for preparing raw data for model training. To have high-quality data to make reliable predictions, the following steps are needed. (i) identifying and handling missing data, (ii) removing duplicates, (iii) handling outliers with unrealistic values of training features. If for one of the data points, all feature training data sets does not exist, that data point is removed. After visually checking the raw dataset and removing the problematic data points, the cleaned training data set with 2037 observation points was used to train the model.\u003c/p\u003e\u003cp\u003eThe suggested database workflow supports a geologically informed and data-consistent training framework, crucial for generating robust and interpretable mineral prospectivity models throughout the Iranian plateau. Implementing ML-based mineral prospectivity mapping needs sophisticated methods for extracting meaningful and geologically plausible patterns from training datasets. A crucial prerequisite for achieving accurate model performance is ensuring compatibility and consistency in the format of training and the target features. The training features comprise diverse geoscientific datasets, each providing complementary insights into mineral systems. The widely used training features are geological data (e.g., Carranza, \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2009\u003c/span\u003e; Li et al., \u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e2021\u003c/span\u003e), geophysical data (e.g., Chen et al., \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2015\u003c/span\u003e; Anderson et al., \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2017\u003c/span\u003e), geochemical data (e.g., Soloviev et al., \u003cspan citationid=\"CR104\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Xiong and Zuo, \u003cspan citationid=\"CR123\" class=\"CitationRef\"\u003e2020\u003c/span\u003e), and remote sensing data (Bruzzone et al., \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2006\u003c/span\u003e; Beygi et al., \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Shirmard et al., \u003cspan citationid=\"CR102\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). In the current study, the training features are geophysical measurements (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e; Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e), mapped geological units (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e; Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e), and spatial variation of the major lithospheric structures (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e; Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). The selected training feature datasets (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e) are underpinned by well-established geological principles and empirical evidence from the Iranian plateau\u0026rsquo;s mineral systems. Each variable serves as a proxy for ore-controlling geological processes. For instance, proximity to igneous rock units indicates potential for porphyry Cu\u0026ndash;Mo\u0026ndash;Au and skarn-type Fe mineralization, especially within the Urumieh-Dokhtar magmatic arc, where Tertiary intrusions have been genetically linked to major deposits such as Sarcheshmeh and Meiduk (Aghazadeh et al., \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2015\u003c/span\u003e). Similarly, specific lithological units such as rhyolite and dacite are associated with high-sulfidation epithermal gold systems and VMS deposits due to their arc volcanic origins. Mafic-ultramafic rocks (e.g., basalt, gabbro) present in ophiolitic belts act as host rocks for Cyprus-type VMS Cu\u0026ndash;Zn deposits, like those found in SE Iran (Hajsadeghi et al., \u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). Structural features such as faults and fault density layers are included due to their role as conduits for hydrothermal fluids and their spatial association with Au and Cu deposits, particularly in the Sanandaj\u0026ndash;Sirjan Zone (Aliyari et al., \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2012\u003c/span\u003e). Moreover, stratigraphic layers reflecting Paleozoic and Mesozoic age units capture the metallogenic potential of formations that host Sedex Pb\u0026ndash;Zn and MVT deposits (e.g., Koushk, Chahmir). These variables collectively reflect the regional metallotectonic framework shaped by Neo-Tethyan subduction and subsequent orogenesis, offering a geologically sound basis for mineral potential modelling. The geophysical data, including aeromagnetic and gravity datasets, are generally provided as continuous raster grids. These raster grids are composed of equally spaced square grids (pixels) and require minimal pre-processing for format alignment. For this study, all geophysical features were harmonized to a raster grid with a 2 km cell size, representing the resolution of spatial analysis. It is expected that the gravity and magnetic field and their derivatives highlight the effects of crustal scale faults and lineaments on the spatial distribution of ore deposits.\u003c/p\u003e\u003cp\u003eThe geological features are provided primarily in vector formats of point data (e.g., ore deposits), polylines (e.g., faults and tectonic boundaries), and polygons (e.g., lithological units). Converting these to raster format enhances the visualization of their spatial relationships. This raster grid allocates each node or pixel the minimum distance to the nearest geological unit, facilitating a uniform and spatially obvious analysis. The transformation was implemented by calculating the Euclidean distance from each raster cell to the closest geological feature. The Euclidean distance is a standard metric for distance estimation of features in the GIS platforms, especially for MPM approaches. Thereby, the value of each point in the transformed raster grids represents the spatial proximity as a continuous variable throughout the grid.\u003c/p\u003e\u003cp\u003eThe distribution of known mineral deposits across the study area is spatially irregular and sparse, often with a lack of comprehensive negative (non-mineralized) training examples. This heterogeneity and discontinuous dataset pose challenges for supervised ML approaches. To address this issue, a spatial density grid estimation of mineral deposits was applied to transform the discrete mineral deposit occurrence data into a continuous surface representing metallogenic intensity. The resulting spatial density grid has a raster grid format with a 2 km cell size to align with the training target.\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eList of the geological and geophysical training feature layers.\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"2\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003eVector data (MD* raster grids), (This study)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eRaster grids\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eIgneous rock units (IG_RU)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e\u003cb\u003eMagnetic anomaly \u0026amp; derivatives\u003c/b\u003e:\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eOphiolite\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eTotal magnetic intensity (TMI; Teknik et al, 2017)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eFault\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eVariable reduced-to-pole (VRTP) TMI, (This study)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eRhyolite\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e1st Vertical gradient of TMI VRTP, (This study)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eGabbro\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e1st x_horizontal gradient of TMI VRTP, (This study)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eGranite\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e1st y_horizontal gradient of TMI VRTP, (This study)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eDiabase\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eAnalytical signal of TMI VRTP, (This study)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eDiorite\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eACMS**\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eBasalt\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e\u003cb\u003eGravity anomalies\u003c/b\u003e (Pavlis et al., 2012):\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eDacite\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eFree Air gravity anomaly (FAA)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eCretaceous\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eAnalytical signal of FAA, (This study)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eDevonian\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eBouguer gravity anomaly (BG)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eEarly Cretaceous\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eAnalytical signal of BG, (This study)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eEarly Jurassic\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e\u003cb\u003eGravity gradients components\u003c/b\u003e ***\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eEarly-Middle Triassic\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eGxx\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eEarly Paleozoic\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eGxy\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eEocene\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eGxz\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eEocene-Oligocene\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eGyy\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eLate Cretaceous\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eGyz\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eCretaceous-Early Paleocene\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eGzz\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eLate Eocene-Oligocene\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e\u003cb\u003eElevation model\u003c/b\u003e (Amante and Eakins, \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2009\u003c/span\u003e)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eLate Jurassic\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eDigital elevation model (DEM)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eLate Jurassic-Early Cretaceous\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e1st Vertical gradient of DEM, (This study)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eLate Paleozoic\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e1st x_horizontal gradient of DEM, (This study)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMiddle Jurassic\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e1st y_horizontal gradient of DEM, (This study)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMiocene\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eAnalytical signal of DEM, (This study)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMiocene-Pliocene\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e\u003cb\u003eLithospheric major structures\u003c/b\u003e \u003cem\u003e(\u003c/em\u003eIrandoust et al., \u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e2022\u003c/span\u003e):\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eOligocene\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eAverage Vs**** of upper crust, (This study)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eOligocene-Miocene\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eAverage Vs**** of lower crust, (This study)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eOrdovician-Silurian\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eAverage Vs**** of upper mantle, (This study)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePaleocene\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eDepth to Moho\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePaleocene-Eocene\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e\u003cb\u003eMiscellaneous\u003c/b\u003e:\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePermian\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eEasting coordinate\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePliocene\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eNorthing coordinate\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePliocene-Quaternary\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eEarthquakes spatial density*****, (This study)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePrecambrian\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eFault lines (Sahandi and Soheili 2014) density\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eQuaternary\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eDensity of boundaries of geological units, (This study)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colspan=\"2\" nameend=\"c2\" namest=\"c1\"\u003e\u003cp\u003e\u003cem\u003e* Minimum distance (calculated in this study) to the lithological units/ages (Sahandi and Soheili, 2014).\u003c/em\u003e\u003c/p\u003e\u003cp\u003e\u003cem\u003e**Averaged crustal magnetic susceptibility (\u003c/em\u003eTeknik et al., \u003cspan citationid=\"CR118\" class=\"CitationRef\"\u003e2020\u003c/span\u003e).\u003c/p\u003e\u003cp\u003e\u003cem\u003e*** The gravity gradients for the study area are at an elevation of 225\u0026thinsp;km above the Earth\u0026rsquo;s surface. The X-axis points to the north, the Y-axis points west, and the Z-axis points up (\u003c/em\u003eBouman et al., \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2016\u003c/span\u003e).\u003c/p\u003e\u003cp\u003e\u003cem\u003e****Vs indicate shear velocity (\u003c/em\u003eIrandoust et al., \u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e2022\u003c/span\u003e).\u003c/p\u003e\u003cp\u003e\u003cem\u003e***** The location of seismic events (Mw\u0026thinsp;\u0026ge;\u0026thinsp;3; from 1940) (\u003c/em\u003e\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://www.isc.ac.uk/isc-ehb/search/catalogue\u003c/span\u003e\u003cspan address=\"http://www.isc.ac.uk/isc-ehb/search/catalogue\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003cem\u003e).\u003c/em\u003e\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003e\u003c/p\u003e"},{"header":"5. Methods","content":"\u003cp\u003eIn this study, ML-based methodologies were implemented on the diverse geospatial datasets for predicting metallogenic potential zones across the Iranian plateau. The workflow commenced with the compilation and harmonization of multi-source geospatial data, including lithological, structural, geophysical, and geochemical layers. These datasets were preprocessed to ensure spatial coordinate consistency. Then, each layer of the dataset is structured into a comprehensive geospatial database to be suitable for supervised learning. For the construction of a structured database, an initial feature selection process was conducted to identify the most relevant training features for metallogenic systems by employing the Minimum Redundancy and Maximum Relevance (MRMR) algorithm (Ding and Peng, \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e2005\u003c/span\u003e). This process utilized statistical analysis and regression-based filtering methods to reduce dimensionality and the effects of the noisy features, essential for improving the efficiency of subsequent modeling processes. A variety of machine learning algorithms were employed to determine the most effective method for predicting areas with high metallogenic potential. Hyperparameters have been optimized for each algorithm through grid-search methods and validation feedback to enhance predictive accuracy. Hyperparameters are typically configuration variables that should be regulated manually to manage machine learning model before training. The hyperparameters are coefficient or mathematical functions that manage the layer number and the size of a neural network, for instance. The hyperparameters of the algorithms (e.g., the learning rate and the batch size) optimize the model learning processes from the data. The trained model on the selected geospatial features, was subsequently applied across the study area, enabling the generation of comprehensive mineral prospectivity maps.\u003c/p\u003e\u003cp\u003eIn this study, we used three distinct datasets including a training dataset, 5-fold cross-validation, and an independent test set (Table S3). The training dataset consisted of 2037 spatial samples each associated with 69 geospatial features derived from geological, geophysical, and lithospheric features as listed in Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e and described in Section \u003cspan refid=\"Sec4\" class=\"InternalRef\"\u003e4\u003c/span\u003e. This dataset was used to fit the regression models. Model performance was validated by 5-fold cross-validation procedure on the training dataset. This technique involved partitioning the 2037 samples of the training dataset into five subsets. Four subsets were used iteratively for training and one was reserved for validation in each iteration, thereby ensuring robustness and reducing overfitting of the models. Model performance was averaged across all folds and evaluated using RMSE and R\u0026sup2; metrics. Additionally, we have an independent test set of 70 well-documented major ore deposits (listed in Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e), which was used to qualitatively assess the correctness of our models. The 70 major ore deposits were not involved in the training or validation processes and were only used to evaluate how well the final model predicted the location of known big metallic mineral. This separation helps us to check how well the model works both on the data it was trained with and on the coordinate of unseen data.\u003c/p\u003e\u003cp\u003eAdditionally, correlation and misfit metrics between the predicted and true occurrence of metallogenic intensity were conducted to assess model reliability. For the models with suboptimal performance (e.g., high RMS or low R\u003csup\u003e2\u003c/sup\u003e values), additional tuning was undertaken. This included removing outliers, refining hyperparameters, modifying feature selection criteria, and adjusting the optimization strategy to improve prediction accuracy and reduce model bias (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eSeven distinct ML algorithms with their 24 subtypes (Table S1), available in the MATLAB regression learner toolbox, were selected to represent a diverse range of algorithm families including tree-based, kernel-based, probabilistic, and neural network methods. The aim is to select an algorithm with optimal performance, suitable for MPM in the Iranian plateau. These major algorithms include: (i) Linear Regression, (ii) Decision Tree, (iii) Vector Machine (SVM), (iv) Ensemble, (v) Gaussian Process Regression, (vi) Neural Network, and (vii) Kernel. Among these, Ensemble-based learning Methods, particularly those employing bagging (bootstrap aggregation), exhibited the most consistent and robust performance.\u003c/p\u003e\u003cp\u003eEnsemble learning methods integrate predictions from multiple base models to improve the overall prediction of the model by enhancing model accuracy and reducing variance. In this study, the Bagging ensemble method uses several base models on bootstrapped subsets of the training data in parallel. The base models are primarily neural network models. The final output was derived by aggregating the most frequent or average individual predictions of the base models (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e). The design and execution of this methodology were successfully used in the recent prospecting and exploration studies (e.g., Yin and Li, \u003cspan citationid=\"CR129\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Chen and Chen, \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; He et al., \u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). The application of the Bagging Ensemble method in this study not only enhanced predictive stability but also showcased its effectiveness in integrating geospatial and geophysical data for mineral prospectivity mapping (MPM).\u003c/p\u003e\u003cp\u003eOnce the optimum model by tuning the hyperparameters is achieved, the trained model used to make predictions over regular grid nodes with a spatial resolution of 2 km \u0026times; 2 km across the Iranian plateau. The cell size is selected based on the average resolution of the geophysical and geological raster grids. The gridded dataset has the same coverage as the aeromagnetic grid of Iran, which covers most of the plateau except for some gaps in the Zagros and Kopeh-Dagh regions (Teknik and Ghods, \u003cspan citationid=\"CR117\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Teknik et al., \u003cspan citationid=\"CR118\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). We cast all datasets into a grid and name those as gridded datasets. It is important to note that this gridded dataset differs from the training feature dataset, which only includes sampling points located at known mineral occurrences. The training dataset consists of 2037 sampling points, while the gridded dataset consists of 377200 sampling points or grid nodes as summarized in Table S3. All the raster grids and rasterized grid layers were then resampled and aligned to the grid coordinate system, ensuring spatial and projection consistency with the training features and target variable, used for the final model prediction.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e"},{"header":"6. Results","content":"\u003cp\u003eAn ML-based approach was developed to predict the spatial variation of metallogenic intensity. To achieve this, we integrated the spatial information of Fe, Cu, Pb, Zn, Au, Ag, and Mg into the metallogenic intensity dataset serving as a training target (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eb). This approach incorporated a comprehensive set of geophysical measurements and geological features (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e) as training datasets. The training datasets comprised 2,037 observation samples. Model performance was assessed through 5-fold cross-validation. Furthermore the validation is qualitatively assessed by using an independent observation point consisting of approximately 70 known metallic deposit locations. Model training was executed on MATLAB's machine learning toolbox using a system equipped with a 7-core Intel i7 GPU @ 2.8 GHz with 64 GB RAM. The training time is not equal for all models. The training models like Ensemble Bagged Trees and Neural Networks require more training time compared to simpler models like Linear Regression. The training was performed by using the parallel computing toolbox of MATLAB R2023b. The cross-validation and hyperparameter tuning were parallelized within loops and model-level parallel execution, which significantly reduced training time in the order of 3\u0026ndash;5 hours, depending on hyperparameters variations.\u003c/p\u003e\u003cp\u003eAn analysis of the training features' importance was conducted to evaluate the contribution of each training feature. Features with high importance alongside qualitative geological studies can enhance the accuracy of the prospective mapping. Moreover, it provides insights for the development of targeted exploration plans in the future as well as deepens our understanding of regional-scale ore formation processes. The Minimum Redundancy and Maximum Relevance (MRMR) algorithm (Ding and Peng, \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e2005\u003c/span\u003e) was used for feature importance analysis. The MRMR algorithm assesses the importance of each training feature on the stable prediction of the model, which is spatial mineralization density in this study. It works by reducing the redundancy among the training features while ensuring their high relevance to the prediction parameter. To achieve this, the mutual information is used to evaluate the features interaction with each other and their relevance to the prediction parameter of the model. Through the importance evaluation process, it removes each training feature from the training process and then estimates the accuracy parameters (e.g., R\u003csup\u003e2\u003c/sup\u003e and RMS) of the trained model. Reduction of the prediction accuracy of the trained model indicates the high importance of the removed training feature. The distribution of importance of the training features shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e indicates that geological (e.g., density of fault lines) and some geophysical (e.g., magnetic and its derivatives) features exhibit high importance. The lithospheric features (e.g., Moho and seismic velocity variations) together with gravity anomalies have lower importance, likely due to their low spatial resolution.\u003c/p\u003e\u003cp\u003eAmong the seven distinct ML algorithms with their 24 subtypes (Table S1), the Ensemble Bagged Trees algorithm demonstrated superior predictive accuracy, achieved by the highest coefficient of determination (R\u0026sup2;) of approximately 0.98 and the lowest root mean square error (RMSE) of about 0.03 (Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). The estimated mineral prospectivity map (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003e) illustrates the spatial density of the mineral occurrence across the Iranian plateau, achieved by Ensemble Bagged Trees algorithm. The high correlation between predicted and the observed values (R\u0026sup2;\u0026asymp;0.98) underscores the robust performance of the supervised Ensemble Bagged Trees algorithm (Fig.\u0026nbsp;\u003cspan refid=\"Fig9\" class=\"InternalRef\"\u003e9\u003c/span\u003e). The Bagged Trees algorithm is very similar to the Random Forest algorithms. Random Forest algorithms are successfully used for MPM (Carranza and Laborte, \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2015\u003c/span\u003e; Parsa and Maghsoudi, \u003cspan citationid=\"CR90\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Rodriguez-Galiano, et al. \u003cspan citationid=\"CR99\" class=\"CitationRef\"\u003e2015\u003c/span\u003e; Yang, et al \u003cspan citationid=\"CR126\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). Therefore, it is not a surprise that we got the highest performance using Bagged Trees methods (Table S1 in the supplementary).\u003c/p\u003e\u003cp\u003eThis regional-scale prospectivity map (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003e) effectively identifies zones with known ore occurrences and predicts underexplored metallogenic zones that are primarily located in northwest, west, and Central Iran (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003e). These prospective zones are predominantly associated with large-scale faults or major tectonic boundaries as well as magmatic complexes. Notably, the results indicate a southward extension of the high metallogenic potential of KaraDagh copper zone, in the northwest of the Iranian plateau. The southward extension is bounded by the NW-SE trending North Tabriz fault. Despite the limited number of training points, the model predicts spatial extension of the Malayer-Isfahan lead-zinc mineral zone toward the west, while a lower spatial density is anticipated eastward along the Kashan-Natanz metallogenic zone (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003e). Additionally, areas nearby the central part of the Zagros suture zone located between Kermanshah and Neyriz ophiolites have been identified as potential targets for further exploration.\u003c/p\u003e\u003cp\u003eTo highlight the underexplored metallogenic zones, a residual density map (Fig.\u0026nbsp;\u003cspan refid=\"Fig17\" class=\"InternalRef\"\u003e10\u003c/span\u003e) was calculated by subtracting the observed spatial mineralization density (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eb) from the predicted density of mineralization density (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003e). This approach facilitates the identification of areas with high metallogenic potential that have not yet been thoroughly investigated. Interestingly, this residual prediction identifies three significant potential metallogenic zones, indicating new prospective areas in the northwest, western, and northern parts of the Iranian plateau.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab3\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable S1\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eThe results are achieved for ML methods. RMSE is the root mean squared error on the validation set. The metrics of R-squared (R2) indicates how well the predicted results explain the target of training. MAE is the Mean absolute error. The MAE is always positive and similar to the RMSE, but less sensitive to outliers. MSE represents mean squared error. The list is sorted in ascending order of RMSE and descending order of R-squared values.\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"6\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMethods Type\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eSubtype\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003eRMSE (Validation)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003eMSE (Validation)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c5\"\u003e\u003cp\u003eRSquared (Validation)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c6\"\u003e\u003cp\u003eMAE (Validation)\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eEnsemble\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eBagged Trees\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.03125\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.00100\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.97620\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e0.01809\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eGaussian Process Regression\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eMatem 5/2 GPR\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.03272\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.00107\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.97245\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e0.01818\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eGaussian Process Regression\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eRational Quadratic GPR\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.03286\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.00108\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.97221\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e0.01821\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eGaussian Process Regression\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eSquared Exponential\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.03337\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.00111\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.97133\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e0.01850\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eNeural Network\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eBi-layered NN\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.03637\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.00132\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.96595\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e0.02410\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eNeural Network\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eTri-layered NN\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.03979\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.00158\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.95924\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e0.02600\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eGaussian Process Regression\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eExponential\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.04054\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.00164\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.95770\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e0.02379\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eNeural Network\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eMedium NN\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.04410\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.00194\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.94995\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e0.02773\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSVM\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eCubic\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.04447\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.00198\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.94909\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e0.02742\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eKernel\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eSVM\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.04651\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.00216\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.94432\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e0.02977\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eNeural Network\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eWide NN\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.04832\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.00234\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.93989\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e0.02886\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSVM\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eQuadratic\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.04948\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.00245\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.93699\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e0.03293\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eNeural Network\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eNarrow\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.05437\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.00296\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.92391\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e0.03477\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSVM\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eMedium Gaussian\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.05508\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.00303\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.92190\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e0.03575\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eTree\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eFine Tree\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.05723\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.00328\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.91569\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e0.03054\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eKernel\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eLeast Squares Regression kernel\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.05964\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.00356\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.90843\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e0.04204\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eTree\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eMedium\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.06426\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.00413\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.89369\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e0.03781\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eEnsemble\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eBoosted Trees\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.06535\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.00427\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.89008\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e0.05057\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eTree\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eCoarse Tree\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.08223\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.00676\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.82594\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e0.05707\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSVM\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eFine Gaussian\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.09723\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.00945\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.75665\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e0.05928\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eLinear Regression\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eLinear\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.11999\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.01440\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.62938\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e0.09175\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSVM\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eLinear\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.15443\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.02385\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.38612\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e0.09373\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSVM\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eCoarse\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.17181\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.02952\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.24011\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e0.10043\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eLinear Regression\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eRobust\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.17434\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.03039\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.21760\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e0.09108\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003ctfoot\u003e\u003ctr\u003e\u003ctd colspan=\"6\"\u003e.\u003c/td\u003e\u003c/tr\u003e\u003c/tfoot\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab4\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable S2\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eSummary of ML Model Hyperparameters and Tuning Ranges\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"4\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003eModel Type\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eSubtype\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003eKey Hyperparameters\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003eTuning Ranges / Notes\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eEnsemble (Bagged Trees)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eBagged Trees\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eNumLearningCycles, MinLeafSize\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eNumLearningCycles: 30; MinLeafSize: 8\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eGaussian Process\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eMatern 5/2 Kernel\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eKernelScale, Sigma, BaiscFunction,\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eKernelScale: auto (fitrgp default), Sigma: auto, BasicFunction: constant\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eNeural Network\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eTri-layered Feedforward\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eNumHiddenLayers, NeuronsPerLayer, LearningRate\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eLayers: 3 (fixed); Neurons: 10\u0026ndash;100 per layer; LearningRate: 0.001\u0026ndash;0.05\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSVM\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eCubic\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eBoxConstraint, KernelScale\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eBoxConstraint: 0.1\u0026ndash;100; KernelScale: 0.01\u0026ndash;10\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eKernel Regression\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eSVM Kernel\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eKernelFunction, Regularization, andIterationLimit\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eKernelFunction: 'gaussian'; Regularization: 0.001\u0026ndash;1; IterationLimit: 1000\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eDecision Tree\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eFine Tree\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eSurrogate decision splits, MinLeafSize\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eSurrogate decision splits: off; MinLeafSize : 1\u0026ndash;5 (optimized at 4)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eLinear Regression\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eLinear\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eNone (least-squares solution)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eRobust option: on\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colspan=\"4\" nameend=\"c4\" namest=\"c1\"\u003e\u003cp\u003eNote: All the models were trained using MATLAB R2023b functions with custom tuning via grid search. The final parameters were selected using 5-fold cross-validation for minimizing RMSE.\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab5\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eThe higher metrics results are achieved for the selected seven different ML methods. The metrics of root mean squared error (RMSE) and R-squared (R2) indicate how well the predicted results explain the target of training. The list is sorted in ascending order of RMSE and, simultaneously, in descending order of R-squared values. The training point number of observations is 2037 with 69 training features or predictors. Validation of the training results is tested with the 5-fold cross-validation. The optimum result is achieved with the Ensemble Bagged Trees method. For details of the training metrics results see Table S1. The best predicted spatial density of the ore occurrences across the model achieved by the Ensemble Bagged Trees method is shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003e. The predictive model for the six other models is presented in Figures S1 to S6 in the supplementary.\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"5\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003eNum.\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eMethod\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003eSub Type\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003eRMSE (Validation)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c5\"\u003e\u003cp\u003eR_Squared\u003c/p\u003e\u003cp\u003e(Validation)\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e1\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e\u003cb\u003eEnsemble\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e\u003cb\u003eBagged Trees\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.031\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.98\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e2\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eGaussian Process Regression\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eMatem 5/2\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.033\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.97\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e3\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eNeural Network\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eTrilayered\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.040\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.96\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003e4\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eSVM\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eCubic\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.044\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.95\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e5\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003ekernel\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eSVM\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.046\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.94\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e6\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eTree\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eFine\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.057\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.91\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e7\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eLinear Regression\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eLinear\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.119\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.62\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab6\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable S3\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eSummary of Data Flow\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"4\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003eDataset\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eSize\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003ePurpose\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003eUsed For\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eTraining Dataset\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e2037\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eModel learning\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eRegression modeling\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eValidation (Cross validation folding)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e5 subsets of training data\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eHyperparameter tuning \u0026amp; model selection\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eRMSE/R\u0026sup2; evaluation\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eIndependent Test Set\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e70\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eExternal model validation\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eFinal qualitatively asses of the trained model\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eGridded dataset\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e377200\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eApply the trained model on the nodes of the grid\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003ePrediction map across the study area with a 2 km \u0026times; 2 km.\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003e\u003c/p\u003e"},{"header":"7. Discussion","content":"\u003cdiv id=\"Sec8\" class=\"Section2\"\u003e\u003ch2\u003e7.1. The mineral prospectivity mapping algorithmic perspective\u003c/h2\u003e\u003cp\u003eThe data-driven approach of mineral prospectivity mapping (MPM) introduces regional-scale exploration solutions by integrating multidimensional geospatial dataset. This geospatial dataset typically consists of various geological features (e.g., rock units and fault lines), lithospheric parameters (e.g., Moho depth and seismic velocity heterogeneity of different layers of the lithosphere), and geophysical measurements (e.g., gravity, magnetic, and their respective derivatives). All of those heterogenous features merged into a unified training dataset. A set of supervised regression-based algorithms are used to derive a predictive model for delineating high-potential metallogenic zones across the Iranian plateau. Using seven major ML algorithms with their 24 subtypes of MATLAB regression learner toolbox, the Ensemble Bagged Trees algorithm performed more accurate and stable predictions for mineral prospectivity across the Iranian plateau, as evidenced with the lowest validation RMSE (~\u0026thinsp;0.03), highest R\u0026sup2; (~\u0026thinsp;0.98) on other metrics (See details at Table S1 in the supplementary). The algorithm inherently has abilities of variance reduction, robustness to overfitting, and handling high-dimensional non-linear geospatial data (e.g., Yin and Li, \u003cspan citationid=\"CR129\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Chen and Chen, \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; He et al., \u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eBagging or bootstrap aggregation lunches multiple decision trees using random subsets of the training data set and then aggregates their outputs to one solution. This approach reduces the potential high variance in the prediction of decision trees by averaging the predictions of all trees. Therefore, using this technique leads to a robust prediction model that is less sensitive to outlier prediction. In our study, the 69 training features represents complex and diverse geological, geophysical, and lithospheric ore-forming controlling parameters. This kind of multidimensional and often spatially noisy data can lead to overfitting in the individual learners. The Bagged Trees method enhances model stability by capturing the dominant prediction without being highly influenced by localized anomalies. The model's ability to derive the non-linear relationships between the training target of ore deposits spatial density and training features is essential for accurate prediction. The strong performance was observed in the cross-validation results. Moreover, the high spatial correlation of the independent 70 major deposit with areas of high values of the predicted density of mineralization, qualitatively provided extra assessment of the performance of the employed model.\u003c/p\u003e\u003cp\u003eThe Ensemble Bagged Trees algorithm shows the relatively accurate prediction of mineralization density, which is important for real-world mineral prospecting mapping. Importantly, this method aligned well with the goals of our study to detect underexplored metallogenic zones. The algorithm\u0026rsquo;s predictive strengths helped us to delineate new potential zones across a geologically complex and tectonically active region of Iranian plateau with diverse types of metallogenic zones, which made the prediction geologically interpretable and highly suitable for regional-scale mineral prospectivity mapping.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec9\" class=\"Section2\"\u003e\u003ch2\u003e7.2. Limitations and challenges of ML-driven mineral prospective mapping\u003c/h2\u003e\u003cp\u003eAlthough the employed ML-driven prospective mapping enhances exploration efficiency in detecting high-potential metallogenic zones, it is essential to recognize its limitations to ensure accurate interpretation. The predictive accuracy of these models is fundamentally dependent on the quality, resolution, and temporal relevance of both training and target datasets (Farahnakian et al., \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). For instance, the dataset of ore deposits used in this study is limited to the most recent (2017 and later) discovered ore deposits. Consequently, the results should be interpreted cautiously, with an awareness of potential gaps or outdated information that may affect the accuracy and comprehensiveness of the analysis (Singer, \u003cspan citationid=\"CR103\" class=\"CitationRef\"\u003e2007\u003c/span\u003e; Mateus and Martins, \u003cspan citationid=\"CR74\" class=\"CitationRef\"\u003e2019\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eIt is crucial to integrate all deposit types as a unique training target data set; otherwise, prediction by individual AI predictive models would be impossible. By considering metallogenic intensity (spatial density of the ore deposits) parameter, we made a unified model of the diverse types of the seven selected ore deposits. Our approach treats the spatial density of all types of ore deposits as a proxy for metallogenic intensity parameter. This means that metallogenic intensity or spatial density of metallogeny across the plateau, structurally and lithologically associated with the same crustal-scale processes (e.g., Sun et al., \u003cspan citationid=\"CR112\" class=\"CitationRef\"\u003e2020\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eWhile different ore deposit types (e.g., porphyry, skarn, vein, SEDEX, MVT) are genetically distinct, but they may still occur within shared mineralisation systems that are governed by common lithospheric-scale processes, including magmatism-related sources, faulting, fluid flow pathways, and depositional environments (McCuaig and Hronsky, \u003cspan citationid=\"CR76\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). Following the mineral systems approach (e.g., Hronsky and Groves, \u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e2008\u003c/span\u003e; McCuaig et al., \u003cspan citationid=\"CR75\" class=\"CitationRef\"\u003e2010\u003c/span\u003e), we argue that these different ore deposits can be regionally associated as an unique parameter, due to their formation within large-scale crustal architectures that condition the source, transport, and focusing of mineralizing fluids. By considering this argument, the similar evidences of different types of ore deposits can be tracked in the geological, geophysical and geochemical observations (Rodriguez-Galiano et al., \u003cspan citationid=\"CR99\" class=\"CitationRef\"\u003e2015\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eRecent mineral system analysis is gradually accepting this approach in ore genesis studies. However, the conceptual framework of linking the mineral system to available data sets is still under development (McCuaig et al., \u003cspan citationid=\"CR75\" class=\"CitationRef\"\u003e2010\u003c/span\u003e; Tagwai et al., \u003cspan citationid=\"CR113\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). While the mineral exploration studies have recognised the approach of this study as useful for the for prospecting metallogenic zones by investigation of ore formation processes (Knox-Robinson and Wyborn, \u003cspan citationid=\"CR62\" class=\"CitationRef\"\u003e1997\u003c/span\u003e; Groves et al., \u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Tagwai et al., \u003cspan citationid=\"CR113\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), using the spatial density as a proxy for mineralization intensity is not intended to imply genetic uniformity, but rather to model the cumulative metallogenic potential imparted by tectonomagmatic controls. This strategy enables us to capture regional mineralization patterns without violating the theoretical distinctions between deposit types. Our results support this view, as spatially clustered mineralization zones align with known trans-lithospheric faults, magmatic arcs, and high-relief tectonic boundaries\u0026mdash;structural features widely recognized as metallogenic drivers.\u003c/p\u003e\u003cp\u003eDespite these limitations, the MPM approach offers an organized approach that prioritizes data-driven targets, minimizes expenses, and enhances discovery rates. To prevent over-interpretation, it is essential to acknowledge the limitations of the input datasets. Integrating advanced ML techniques with high-quality updated geospatial data enables MPM to optimize exploration strategies and enhance the likelihood of successful ore deposit explorations (Zhang et al., \u003cspan citationid=\"CR134\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eA central challenge in this study arises from the irregular distribution of the discovered ore deposits (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e), which underscores both the geological complexity of the Iranian plateau and the influence of historically biased exploration efforts (Lou and Liu, \u003cspan citationid=\"CR69\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). The observed spatial pattern, marked by clustering and sparsity, highlights the challenges of directly attributing raster grid training features to discrete deposit locations due to the variable intensity of the mineralization. The irregular spatial distribution of the ore deposits can cause less uncertain prediction, especially where the distribution is sparse or different ore deposits are highly clustered. To overcome these limitations and align with advancements in handling imbalanced geospatial data, the spatial density of the mineral deposits is selected as the target variable for ML model, instead of depending on specific mineral deposit coordinates (Farahnakian et al., \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). To evaluate the spatial uncertainty of the predictions, a residual map (Figure S7) was generated, showing the difference between true and predicted spatial densities of ore deposits at known locations. Residuals mostly fall within \u0026plusmn;\u0026thinsp;0.1, suggesting good model performance overall. However, larger residuals (\u0026gt;|\u0026plusmn;0.1|) are observed in under-sampled areas, indicating greater model uncertainty. These residuals are particularly notable for isolated deposits distant from major metallogenic clusters. Interestingly, even in some highly sampled regions, such as the KaraDagh metallogenic zone in the northwest of the Iranian plateau, a localized cluster of elevated residuals is detected, possibly due to local geological complexity.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eRelatively low feature importance scores of lithospheric and gravity features (e.g., Moho, crustal seismic velocity, and Bouguer gravity anomalies) are in contrast with the generally accepted importance of lithospheric structures in controlling metallogenic zones. The Moho depth and Bouguer anomalies have spatially low resolution (\u0026thinsp;~\u0026thinsp;\u0026gt;\u0026thinsp;0.5\u0026deg;). These low-resolution lithospheric features contribute less to localized variance of the metallogenic spatial density (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e). However, in this study qualitatively and less quantitatively, they still provide important geological context by delineating major crustal domains, tectonic boundaries, and zones of magmatic activities, all of which influence ore-forming processes.\u003c/p\u003e\u003cp\u003eIt is worth noting that the ore deposits dataset used here was last updated in 2017. Therefore, any exploration conducted since then could be vital in independently validating or refining our predictions. Therefore, newly identified ore occurrences would provide valuable ground-truth evidence for the model's reliability and help fine-tune prospectivity interpretations.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec10\" class=\"Section2\"\u003e\u003ch2\u003e7.3. Geological implications of the results\u003c/h2\u003e\u003cp\u003eThe suitable condition for the formation of metallic ore deposits is inherently associated with a specific physicochemical condition, geodynamic processes, and tectonic evolution. Therefore, the spatial mineral occurrences are typically attributed to the geological features (e.g., lithological units, fault systems, magmatic intrusions, and hydrothermal fluids). The geophysical anomalies (e.g., gravity and magnetic) reflect those geological structures, depending their sensitivity. These complex ore deposit formation mechanisms within various types of geological structures and their complicated geophysical signatures pose significant challenges to regional metallogenic zone mapping (Chernicoff et al., \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2002\u003c/span\u003e; Bierlein et al., \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2006\u003c/span\u003e; Leclerc et al., \u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e2012\u003c/span\u003e; Dufr\u0026eacute;chou et al., \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e2015\u003c/span\u003e; Jamali and Mehrabi, \u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e2015\u003c/span\u003e; Bauer et al., \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2022\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eThe residual spatial density of the mineral deposits (Fig.\u0026nbsp;\u003cspan refid=\"Fig17\" class=\"InternalRef\"\u003e10\u003c/span\u003e) was derived by subtracting the calculated spatial density of the ore deposits (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ea) from the model\u0026rsquo;s prediction (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003e). The residual density map provides insights to pinpoint areas across the Iranian plateau with high metallogenic potential, which have remained underexplored. The results indicate that the promising areas are closely associated with the known major metallogenic zones (Fig.\u0026nbsp;\u003cspan refid=\"Fig17\" class=\"InternalRef\"\u003e10\u003c/span\u003e). The residual anomalies are spatially aligned with established metallogenic provinces, such as the KaraDagh, Malayer-Isfahan, and Toroud zones (see Figs.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ed and \u003cspan refid=\"Fig17\" class=\"InternalRef\"\u003e10\u003c/span\u003e). This pattern indicates that the model accurately represents the main litho-structural and magmatic-tectonic controls on ore formation in the region. The zones lie within geodynamically active regions shaped by the convergence of the Arabian and Eurasian plates, where processes such as crustal shortening, strike-slip faulting, and arc magmatism have historically facilitated the formation of systems including porphyry Cu-Au-Mo ore deposits. The Central Iran block (e.g., Yazd, Tabas, and Lut blocks) exhibits comparatively weak residual signals, which presumably indicates that significant mineralized belts in this region have been well-mapped by previous explorations. However, some limited, structurally aligned anomalies trace the terrane boundaries and mapped faults. This observation highlights the significance of deep crustal fractures in directing hydrothermal fluids in central Iran. We argue that the distribution of ore deposits is not controlled only by active faults but controlled by all crustal scale fractures, old faults, and inactive lineaments. A better fault and lineament map will affect our results to be more precise but because of using a coarse fault map, the effect of all crustal-scale faults and fractures is missing in our results. However, the topography and magnetic field with their derivatives may provide the expected effect of those crustal-scale cracks, old faults, and inactive lineaments, which are not presented in the fault map of Iran.\u003c/p\u003e\u003cp\u003eEspecially, the high feature importance of the DEM (Digital Elevation Model) layer (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e) and its derivatives indicate its typical correspondence to the tectonically active regions such as fault zones, terrane boundaries, suture zones and magmatic intrusions, which are spatially associated with ore deposit (e.g., porphyry Cu, skarn, epithermal Au) (e.g., Yang et al., \u003cspan citationid=\"CR126\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Zuo and Xu, \u003cspan citationid=\"CR136\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Therefore, the high feature importance of topography does not imply that elevation causes mineralization, but instead the DEM acts as a proxy variable that captures the spatial pattern of geological processes, which are conducive to ore deposit formation. For instance, in the Iranian plateau, the Urumieh-Dokhtar magmatic arc, important for hosting porphyry and skarn deposits, is marked with high relief originating from intense subduction-related magmatism. The prospectivity maps also hold relevance for critical minerals of rare earth elements, which are often genetically related to metallic ore-forming systems (e.g., Petrella et al., \u003cspan citationid=\"CR94\" class=\"CitationRef\"\u003e2014\u003c/span\u003e; Zarasvandi et al., \u003cspan citationid=\"CR132\" class=\"CitationRef\"\u003e2015\u003c/span\u003e; Abedini et al., \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2023\u003c/span\u003e).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec11\" class=\"Section2\"\u003e\u003ch2\u003e7.4. AI-methods potentials for MPM and suggestions for future studies\u003c/h2\u003e\u003cp\u003eThis study indicates the efficiency of applying modern ML algorithms on the various geospatial data for prospecting metallogenic zones on a regional scale, especially in areas with diverse tectonics, such as Iranian plateau. Employing this approach provided valuable regional insights into potential metallogenic zones by including only positive training dataset. Positive training data points indicate observation points with confirmed mineralization, while negative data points represent locations where ore deposit formation is unlikely. The applicability of the AI-driven approaches can be broadened through the integration of advanced deep learning methods by including both positive and negative training datasets. The aim is to provide interpretable AI-aided algorithms in mineral exploration with a promising trajectory in MPM. Enhancing the workflow by using new ML methods alongside advanced hyperparameter optimization would help mitigate overfitting and more effectively manage non-linearity problems between training dataset and the target of the method (Chen and Wu, \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Li et al., \u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e2021\u003c/span\u003e, \u003cspan citationid=\"CR65\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Maepa et al., \u003cspan citationid=\"CR70\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Parsa and Carranza, \u003cspan citationid=\"CR89\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Parsa and Maghsoudi, \u003cspan citationid=\"CR90\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; N. Yang et al., \u003cspan citationid=\"CR126\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Yin et al., \u003cspan citationid=\"CR128\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Zuo and Xu, \u003cspan citationid=\"CR136\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Wake et al., \u003cspan citationid=\"CR121\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eTo achieve a trained model with higher prediction accuracy, it is essential to incorporate diverse geospatial datasets with higher resolution. For instance, integration of radiometric and geochemical sampling together with satellite-driven multispectral or hyperspectral imagery can significantly improve the prediction accuracy of the trained model. These additional datasets, especially hyperspectral images, identify hydrothermal alteration indicators, geochemical anomalies, and structurally controlled mineralizing systems. Applying this approach will enhance the model sensitivity to characteristics such as iron zones, clay alteration, or silicification, which typically indicative of deposit types like porphyries, epithermal systems, or skarns.\u003c/p\u003e\u003cp\u003eWhile this study focused on identifying prospective metallogenic zones, regardless of mineralization type, the workflow could be adapted to distinguish deposit types by classification learning algorithms using spectral, geophysical, or geochemical training dataset (Abedi et al., \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2012\u003c/span\u003e; Shirmard et al., \u003cspan citationid=\"CR102\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Chen and Chen, \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Farahnakian et al., \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Mahboob et al., \u003cspan citationid=\"CR73\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). For example, segmentation algorithms could relatively differentiate porphyry copper systems, characterized by halos and disseminated sulfides alteration, from epithermal gold-silver deposits, which are marked by alteration zones of vein-hosted metallogenic and intense silicification (Shayeganpour and Tangestani, \u003cspan citationid=\"CR101\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Amraoui et al., \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2025\u003c/span\u003e; Farahbakhsh et al., \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). Employing such approaches would facilitate more precise targeting of specific deposit types. These approaches, when properly implemented, would yield significant insights into spatial patterns of the ore deposits, thereby enhancing mineral exploration strategies and decision-making processes.\u003c/p\u003e\u003c/div\u003e"},{"header":"8. Conclusions","content":"\u003cp\u003eThe AI-aided mineral prospectivity mapping (MPM) approach is used to identify new metallogenic zones in the Iranian plateau. Analyzing regional-scale geological and geophysical data is typically associated with challenges when using traditional methods. AI algorithms facilitate resolving the nonlinear and complex patterns among irregularly distributed mineral deposits using multimodal geological and geophysical big data. Regression learning methods are used to identify the intrinsic relationship between various geological and geophysical training features and the locations of the metallic ore deposits as training targets.\u003c/p\u003e\u003cp\u003eThe training target dataset is formed by integrating the location of seven metallic ore deposits including Fe, Cu, Pb, Zn, Au, Ag, and Mg. Then the spatial density of ore deposits is computed using the integrated dataset. The information regarding the grade and tonnage of various ore deposits in our dataset is limited, thus identical importance weights are assigned to each ore deposit. To form training feature dataset, the vector-based features (e.g., lithological units and fault lines) were transformed into a set of raster grids to ensure their data format consistency with geophysical and lithospheric raster grids. The raster grids and vector-based features are created by calculating the minimum distance to geological units and lithospheric structural boundaries. These created raster grids together with raster grids of geophysical measurements and mid-lithospheric interfaces (e.g., average crustal shear waves and Moho depth), formed the training dataset comprising 69 features. Feature importance analysis is conducted to identify the most relevant features for ore formation.\u003c/p\u003e\u003cp\u003eThe learning performance is evaluated for seven major regressions-based algorithms with their subtypes reaching 24 methods. The Ensemble Bagged Trees method displayed the best performance. The selection criteria were the minimum RMSE (~\u0026thinsp;0.03) and maximum R-Squared (R\u003csup\u003e2\u003c/sup\u003e\u0026thinsp;\u0026asymp;\u0026thinsp;0.98) value from the regression analysis of prediction versus observation of validation data. Once the model is trained and its accuracy adheres to the acceptable critical, it is employed to estimate the spatial density of the ore occurrences at grid nodes with a cell spacing of 2 km across the Iranian plateau. The predicted spatial density of the ore deposits is compared with locations of comprehensive data on world-class ore deposits. To highlight underexplored areas, residual spatial density anomalies were calculated by subtracting the observed from the predicted spatial density. The results indicate the prediction is optimally aligned with the spatial association of major ore deposits. The trends of predicted spatial density indicates a new zone of metallogenic activity close to the existing metallogenic zones. This study suggests that the structural boundaries and magmatic complexes along with geophysical observations, are significant indicators of metallic ore deposits and should be prioritized in future mineral exploration plans.\u003c/p\u003e\u003cp\u003eThe results indicate that the high-potential metallogenic zones are primarily located in the northwest of the study area, aligning closely with the known distribution of the Copper metallogenic zone of the KaraDagh. The southern boundary of this zone is primarily outlined by the major North Tabriz fault. The predicted area on the western side of the Zagros suture aligns closely with the known Malayer-Isfahan Pb-Zn metallogenic zone. The residual spatial density of the ore deposits suggests that Central Iran has already been almost fully explored, revealing fewer unexplored metallogenic zones. Nonetheless, a locally less explored metallogenic zone has been identified near the Bafgh, Nehbandan-Ferdous, and Jiroft-Shaherbabak metallogenic zones.\u003c/p\u003e\u003cp\u003eThe results of this study are useful for enhancing local-scale exploration success by employing a practical and multidisciplinary workflow of reginal-scale mineral prospective mapping (MPM). This approach addresses the rising demands for ore minerals, considering sustainable development and the economic limitations of mineral exploration. Leveraging the recent advancements in the big data analysis and artificial intelligence algorithms aid environmentally sustainable mineral exploration.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003ch2\u003eData/Software availability statement\u003c/h2\u003e\u003cp\u003eThe National Centers for Environmental Information (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.ncei.noaa.gov/products/etopo-global-relief-model\u003c/span\u003e\u003cspan address=\"https://www.ncei.noaa.gov/products/etopo-global-relief-model\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003e)\u003c/span\u003e was used to access the ETOPO1 global elevation model (Amante and Eakins, \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2009\u003c/span\u003e) from the ETOPO Global Relief Model grid. The Cenozoic volcano\u0026rsquo;s locations are downloaded from the Smithsonian Institution's Global Volcanism Program (GVP) website (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://volcano.si.edu/\u003c/span\u003e\u003cspan address=\"https://volcano.si.edu/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003e).\u003c/span\u003e The gravity gradient grids at 225 km and 255 km height are available from the website of the European Space Agency (ESA) (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://earth.esa.int/eogateway/catalog/goce-global-gravity-field-models-and-grids\u003c/span\u003e\u003cspan address=\"https://earth.esa.int/eogateway/catalog/goce-global-gravity-field-models-and-grids\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003e)\u003c/span\u003e (Bouman et al., \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). The EGM2008 global Bouguer anomaly model (Pavlis et al., 2012) has been accessed by the National Geospatial-Intelligence Agency from the website (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://bgi.obs-mip.fr/grids-and-models-2/grids-and-models-2-2/#toc7\u003c/span\u003e\u003cspan address=\"https://bgi.obs-mip.fr/grids-and-models-2/grids-and-models-2-2/#toc7\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003e).\u003c/span\u003e The MATLAB codes are used to prepare data and compute the models that can be provided by the author. The lithological units/ages (Sahandi and Soheili 2014). Averaged crustal magnetic susceptibility (Teknik et al. \u003cspan citationid=\"CR118\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). Vs indicate shear velocity (Irandoust et al., \u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). The location of seismic events (Mw\u0026thinsp;\u0026ge;\u0026thinsp;3; from 1940) (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://www.isc.ac.uk/isc-ehb/search/catalogue\u003c/span\u003e\u003cspan address=\"http://www.isc.ac.uk/isc-ehb/search/catalogue\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e).\u003c/p\u003e\u003c/p\u003e\u003cp\u003e\u003ch2\u003eConflict of Interest\u003c/h2\u003e\u003cp\u003eThe authors have no conflicts of interest to declare that are relevant to the content of this article\u003c/p\u003e\u003c/p\u003e\u003ch2\u003eAcknowledgments\u003c/h2\u003e\u003cp\u003eThis research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors. Vahid Teknik (VT) extends his special appreciation to Ebrahim Gholamzadeh, whose support made this research possible. VT thanks Irina Artemieva and Hans Thybo for helpful scientific support. Amr Abdelnasser extends his appreciation to the Scientific Research Project (BAP Project ID: 46804) at Istanbul Technical University (ITU, Turkey).\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eAbedi M, Norouzi GH, Bahroudi A (2012) Support vector machine for multi-classification of mineral prospectivity areas. Comput Geosci 46:272\u0026ndash;283. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.cageo.2011.12.014\u003c/span\u003e\u003cspan address=\"10.1016/j.cageo.2011.12.014\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eAbedini M, Ziaii M, Timkin T, Pour AB (2023) Machine Learning (ML)-Based Copper Mineralization Prospectivity Mapping (MPM) Using Mining Geochemistry Method and Remote Sensing Satellite Data. Remote Sens 2023 15(15):3708. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/RS15153708\u003c/span\u003e\u003cspan address=\"10.3390/RS15153708\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eAdiri Z, Lhissou R, El Harti A, Jellouli A, Chakouri M (2020) Recent advances in the use of public domain satellite imagery for mineral exploration: A review of Landsat-8 and Sentinel-2 applications. Ore Geol Rev 117. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.oregeorev.2020.103332\u003c/span\u003e\u003cspan address=\"10.1016/j.oregeorev.2020.103332\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eAghazadeh M, Hou Z, Badrzadeh Z, Zhou L (2015) Temporal-spatial distribution and tectonic setting of porphyry copper deposits in Iran: Constraints from zircon U-Pb and molybdenite Re-Os geochronology. Ore Geol Rev 70:385\u0026ndash;406. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/J.OREGEOREV.2015.03.003\u003c/span\u003e\u003cspan address=\"10.1016/J.OREGEOREV.2015.03.003\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eAlaminia Z, Karimpour MH, Homam SM, Finger F (2013) The magmatic record in the Arghash region (northeast Iran) and tectonic implications. Int J Earth Sci 102:1603\u0026ndash;1625. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s00531-013-0897-1\u003c/span\u003e\u003cspan address=\"10.1007/s00531-013-0897-1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eAlavi M (1996) Tectonostratigraphic synthesis and structural style of the Alborz Mountain System in Iran. J Geodyn 21:1\u0026ndash;33\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eAliyari F, Rastad E, Mohajjel M (2012) Gold Deposits in the Sanandaj-Sirjan Zone: Orogenic Gold Deposits or Intrusion-Related. Gold Systems? Resource Geol 62:296\u0026ndash;315. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/J.1751-3928.2012.00196\u003c/span\u003e\u003cspan address=\"10.1111/J.1751-3928.2012.00196\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.X;PAGEGROUP:STRING:PUBLICATION\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eAllen MB, Kheirkhah M, Neill I, Emami MH, Mcleod CL (2013) Generation of Arc and Within-plate Chemical Signatures in Collision Zone Magmatism: Quaternary Lavas from Kurdistan Province, Iran. J Petrol 54:887\u0026ndash;911. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1093/petrology/egs090\u003c/span\u003e\u003cspan address=\"10.1093/petrology/egs090\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eAmante C, Eakins BW (2009) ETOPO1 1 Arc-Minute Global Relief Model: Procedures, Data Sources and Analysis, NOAA Technical Memorandum NESDIS NGDC-24. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1594/PANGAEA.769615\u003c/span\u003e\u003cspan address=\"10.1594/PANGAEA.769615\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eAmraoui T, Ibouh H, Farah A, Bammou Y, Shebl A (2025) Remote sensing mapping of structural and hydrothermal alteration in the mougueur inlier, Eastern high atlas, Morocco. Sci Rep 15:1\u0026ndash;18. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/S41598-025-99402-0\u003c/span\u003e\u003cspan address=\"10.1038/S41598-025-99402-0\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e;SUBJMETA=213,2151,330,431,704;KWRD=GEOLOGY,MINERALOGY,PETROLOGY\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eAnderson ED, Monecke T, Hitzman MW, Zhou W, Bedrosian PA (2017) Mineral Potential mapping in an accreted island-Arc setting using aeromagnetic data: An example from Southwest Alaska. Econ Geol 112:375\u0026ndash;396. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.2113/ECONGEO.112.2.375\u003c/span\u003e\u003cspan address=\"10.2113/ECONGEO.112.2.375\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eArjmandzadeh R, Karimpour MH, Mazaheri SA, Santos JF, Medina JM, Homam SM (2011) Sr-Nd isotope geochemistry and petrogenesis of the Chah-Shaljami granitoids (Lut Block, Eastern Iran). J Asian Earth Sci 41:283\u0026ndash;296. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/J.JSEAES.2011.02.014\u003c/span\u003e\u003cspan address=\"10.1016/J.JSEAES.2011.02.014\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eAsadi S, Moore F, Zarasvandi A (2014) Discriminating productive and barren porphyry copper deposits in the southeastern part of the central Iranian volcano-plutonic belt, Kerman region, Iran: A review. Earth Sci Rev 138:25\u0026ndash;46. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/J.EARSCIREV.2014.08.001\u003c/span\u003e\u003cspan address=\"10.1016/J.EARSCIREV.2014.08.001\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBagheri H (2015) Crustal lineament control on mineralization in the Anarak area of Central Iran. Ore Geol Rev 66:293\u0026ndash;308. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/J.OREGEOREV.2014.10.028\u003c/span\u003e\u003cspan address=\"10.1016/J.OREGEOREV.2014.10.028\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBauer TE, Lynch EP, Sarlus Z, Drejing-Carroll D, Martinsson O, Metzger N, Wanhainen C (2022) Structural Controls on Iron Oxide Copper-Gold Mineralization and Related Alteration in a Paleoproterozoic Supracrustal Belt: Insights from the Nautanen Deformation Zone and Surroundings, Northern Sweden. Econ Geol 117:327\u0026ndash;359. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.5382/ECONGEO.4862\u003c/span\u003e\u003cspan address=\"10.5382/ECONGEO.4862\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBenaissi L, Tarek A, Tobi A, Ibouh H, Zaid K, Elamari K, Hibti M (2022) Geological mapping and mining prospecting in the Aouli inlier (Eastern Meseta, Morocco) based on remote sensing and geographic information systems (GIS). China Geol 5:614\u0026ndash;625. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.31035/cg2022035\u003c/span\u003e\u003cspan address=\"10.31035/cg2022035\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBerberian M, King GCP (1981) Towards a paleogeography and tectonic evolution of Iran. Can J Earth Sci 18:210\u0026ndash;265. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1139/e81-019\u003c/span\u003e\u003cspan address=\"10.1139/e81-019\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBergen KJ, Johnson PA, de Hoop MV, Beroza GC (2019) Machine learning for data-driven discovery in solid Earth geoscience. Science (1979) 363. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1126/science.aau0323\u003c/span\u003e\u003cspan address=\"10.1126/science.aau0323\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBeygi S, Talovina IV, Tadayon M, Pour AB (2021) Alteration and structural features mapping in Kacho-Mesqal zone, Central Iran using ASTER remote sensing data for porphyry copper exploration. Int J Image Data Fusion 12:155\u0026ndash;175. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/19479832.2020.1838628\u003c/span\u003e\u003cspan address=\"10.1080/19479832.2020.1838628\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBierlein FP, Groves DI, Goldfarb RJ, Dub\u0026eacute; B (2006) Lithospheric controls on the formation of provinces hosting giant orogenic gold deposits. Min Depos 40:874\u0026ndash;886. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/S00126-005-0046-2\u003c/span\u003e\u003cspan address=\"10.1007/S00126-005-0046-2\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBoulin J (1991) Structures in Southwest Asia and evolution of the eastern Tethys. Tectonophysics 196:211\u0026ndash;268. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/0040-1951(91)90325-M\u003c/span\u003e\u003cspan address=\"10.1016/0040-1951(91)90325-M\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBouman J, Ebbing J, Fuchs M, Sebera J, Lieb V, Szwillus W, Haagmans R, Novak P (2016) Satellite gravity gradient grids for geophysics. Sci Rep 6:1\u0026ndash;11. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/srep21050\u003c/span\u003e\u003cspan address=\"10.1038/srep21050\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBruzzone L, Chi M, Marconcini M (2006) A novel transductive SVM for semisupervised classification of remote-sensing images. IEEE Trans Geosci Remote Sens 44:3363\u0026ndash;3373. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1109/tgrs.2006.877950\u003c/span\u003e\u003cspan address=\"10.1109/tgrs.2006.877950\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBurg JP (2018) Geology of the onshore Makran accretionary wedge: Synthesis and tectonic interpretation. Earth Sci Rev 185:1210\u0026ndash;1231. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.earscirev.2018.09.011\u003c/span\u003e\u003cspan address=\"10.1016/j.earscirev.2018.09.011\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eCarranza EJM (2009) Controls on mineral deposit occurrence inferred from analysis of their spatial pattern and spatial association with geological features. Ore Geol Rev 35:383\u0026ndash;400. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.oregeorev.2009.01.001\u003c/span\u003e\u003cspan address=\"10.1016/j.oregeorev.2009.01.001\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eCarranza EJM, Laborte AG (2015) Random forest predictive modeling of mineral prospectivity with small number of prospects and data with missing values in Abra (Philippines). Comput Geosci 74:60\u0026ndash;70. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.cageo.2014.10.004\u003c/span\u003e\u003cspan address=\"10.1016/j.cageo.2014.10.004\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eChen G, Cheng Q, Zuo R, Liu T, Xi Y (2015) Identifying gravity anomalies caused by granitic intrusions in Nanling mineral district, China: a multifractal perspective. Geophys Prospect 63:256\u0026ndash;270. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/1365-2478.12187\u003c/span\u003e\u003cspan address=\"10.1111/1365-2478.12187\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eChen J, Chen Y (2023) A high-performance voting-based ensemble model of graph convolutional extreme learning machines for identifying geochemical anomalies related to mineralization. Ore Geol Rev 162:1\u0026ndash;14. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.oregeorev.2023.105706\u003c/span\u003e\u003cspan address=\"10.1016/j.oregeorev.2023.105706\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eChen Y, Wu W (2017) Mapping mineral prospectivity using an extreme learning machine regression. Ore Geol Rev 80:200\u0026ndash;213. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.oregeorev.2016.06.033\u003c/span\u003e\u003cspan address=\"10.1016/j.oregeorev.2016.06.033\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eChernicoff CJ, Richards JP, Zappettini EO (2002) Crustal lineament control on magmatism and mineralization in northwestern Argentina: Geological, geophysical, and remote sensing evidence. Ore Geol Rev 21:127\u0026ndash;155. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/S0169-1368(02)00087-2\u003c/span\u003e\u003cspan address=\"10.1016/S0169-1368(02)00087-2\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eChiu HY, Chung SL, Zarrinkoub MH, Mohammadi SS, Khatib MM, Iizuka Y (2013) Zircon U-Pb age constraints from Iran on the magmatic evolution related to Neotethyan subduction and Zagros orogeny. Lithos 162\u0026ndash;163:70\u0026ndash;87. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/J.LITHOS.2013.01.006\u003c/span\u003e\u003cspan address=\"10.1016/J.LITHOS.2013.01.006\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eChukwu C, Betts P, Moore D, Munukutla R, Armit R, McLean M, Grose L (2024) Unsupervised machine learning and depth clusters of Euler deconvolution of magnetic data: a new approach to imaging geological structures. Explor Geophys 55:223\u0026ndash;245. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/08123985.2023.2299475\u003c/span\u003e\u003cspan address=\"10.1080/08123985.2023.2299475\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eDaliran F (2008) The carbonate rock-hosted epithermal gold deposit of Agdarreh, Takab geothermal field, NW Iran - Hydrothermal alteration and mineralisation. Min Depos 43:383\u0026ndash;404. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/S00126-007-0167-X\u003c/span\u003e\u003cspan address=\"10.1007/S00126-007-0167-X\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eDavies RS, Davies MJ, Groves D, Davids K, Brymer E, Trench A, Sykes JP, Dentith M (2021) Learning and Expertise in Mineral Exploration Decision-Making: An Ecological Dynamics Perspective. Int J Environ Res Public Health 18:9752. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/ijerph18189752\u003c/span\u003e\u003cspan address=\"10.3390/ijerph18189752\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eDentith M, Enkin RJ, Morris W, Adams C, Bourne B (2020) Petrophysics and mineral exploration: a workflow for data analysis and a new interpretation framework. Geophys Prospect 68:178\u0026ndash;199. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/1365-2478.12882\u003c/span\u003e\u003cspan address=\"10.1111/1365-2478.12882\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eDentith M, Mudge S (2014) Geophysics for the mineral exploration geoscientist. AusIMM Bull. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1017/CBO9781139024358\u003c/span\u003e\u003cspan address=\"10.1017/CBO9781139024358\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eDing C, Peng H (2005) Minimum redundancy feature selection from microarray gene expression data. J Bioinform Comput Biol 3:185\u0026ndash;205\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eDufr\u0026eacute;chou G, Harris LB, Corriveau L, Antonoff V (2015) Regional and local controls on mineralization and pluton emplacement in the Bondy gneiss complex, Grenville Province, Canada interpreted from aeromagnetic and gravity data. J Appl Geophy 116:192\u0026ndash;205. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/J.JAPPGEO.2015.03.015\u003c/span\u003e\u003cspan address=\"10.1016/J.JAPPGEO.2015.03.015\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eEnayat M, Ghods A (2023) 3D Shear-Wave Velocity Model of Central Makran Using Ambient-Noise Adjoint Tomography. J Geophys Res Solid Earth 128. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1029/2023JB026928\u003c/span\u003e\u003cspan address=\"10.1029/2023JB026928\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e;PAGEGROUP:STRING:PUBLICATION e2023JB026928\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eFarahbakhsh E, Goel D, Pimparkar D, Muller RD, Chandra R (2025) Convolutional neural networks for mineral prospecting through alteration mapping with remote sensing data. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s41064-025-00344-z\u003c/span\u003e\u003cspan address=\"10.1007/s41064-025-00344-z\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eFarahnakian F, Sheikh J, Zelioli L, Nidhi D, Sepp\u0026auml; I, Ilo R, Nevalainen P, Heikkonen J (2024) Addressing imbalanced data for machine learning based mineral prospectivity mapping. Ore Geol Rev 174:106270. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/J.OREGEOREV.2024.106270\u003c/span\u003e\u003cspan address=\"10.1016/J.OREGEOREV.2024.106270\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eGeranian H, Tabatabaei SH, Asadi HH, Carranza EJM (2016) Application of Discriminant Analysis and Support Vector Machine in Mapping Gold Potential Areas for Further Drilling in the Sari-Gunay Gold Deposit, NW Iran. Nat Resour Res 25:145\u0026ndash;159. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/S11053-015-9271-2\u003c/span\u003e\u003cspan address=\"10.1007/S11053-015-9271-2\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eGolmohammadi A, Karimpour MH, Shafaroudi AM, Mazaheri SA (2015) Alteration-mineralization, and radiometric ages of the source pluton at the Sangan iron skarn deposit, northeastern Iran. Ore Geol Rev 65:545\u0026ndash;563\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eGhorbani M (2013a) The economic geology of Iran: Mineral deposits and natural resources. The Economic Geology of Iran: Mineral Deposits and Natural Resources 1\u0026ndash;569. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/978-94-007-5625-0\u003c/span\u003e\u003cspan address=\"10.1007/978-94-007-5625-0\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eGhorbani M (2013b) Metallogenic and mining provinces, belts and zones of Iran. Springer Geol 199\u0026ndash;295. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/978-94-007-5625-0_6/FIGURES/26\u003c/span\u003e\u003cspan address=\"10.1007/978-94-007-5625-0_6/FIGURES/26\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eGlennie KW, Hughes Clarke MW, Boeuf MGA, Pilaar WFH, Reinhardt BM (1990) Inter-relationship of Makran-Oman Mountains belts of convergence. Geological Society, London. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1144/GSL.SP.1992.049.01.47\u003c/span\u003e\u003cspan address=\"10.1144/GSL.SP.1992.049.01.47\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e., Special Publications\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eGonzalez-Alvarez I, Goncalves MA, Carranza EJM (2020) Introduction to the Special Issue Challenges for mineral exploration in the 21st century: Targeting mineral deposits under cover. Ore Geol Rev 126:103785. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.oregeorev.2020.103785\u003c/span\u003e\u003cspan address=\"10.1016/j.oregeorev.2020.103785\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eGroves DI, Santosh M, M\u0026uuml;ller D, Zhang L, Deng J, Yang LQ, Wang QF, Mineral systems: Their advantages in terms of developing holistic genetic models and for target generation in global mineral exploration. Geosystems and Geoenvironment 1., Bierlein DI (2022) F.P., 2007. Geodynamic settings of mineral deposit systems. J Geol Soc London 164, 19\u0026ndash;30. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1144/0016-76492006-065\u003c/span\u003e\u003cspan address=\"10.1144/0016-76492006-065\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eGuo P, Yang T (2023) Quantifying Continental Crust Thickness Using the Machine Learning Method. J Geophys Res Solid Earth 128:1\u0026ndash;16. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1029/2022JB025970\u003c/span\u003e\u003cspan address=\"10.1029/2022JB025970\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHajsadeghi S, Mirmohammadi M, Asghari O, Meshkani SA (2018) Geology and mineralization at the copper-rich volcanogenic massive sulfide deposit in Nohkouhi, Posht-e-Badam block, Central Iran. Ore Geol Rev 92:379\u0026ndash;396. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/J.OREGEOREV.2017.11.030\u003c/span\u003e\u003cspan address=\"10.1016/J.OREGEOREV.2017.11.030\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHassanlouei BT, Rajabzadeh MA (2019) Iron ore deposits associated with Hormuz evaporitic series in Hormuz and Pohl salt diapirs, Hormuzgan province, southern Iran. J Asian Earth Sci 172:30\u0026ndash;55. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.jseaes.2018.08.024\u003c/span\u003e\u003cspan address=\"10.1016/j.jseaes.2018.08.024\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHassanpour S, Rajabpour S (2020) Magmatic-hydrothermal evolution of the Anjerd Cu skarn deposit, NW Iran: perspectives on mineral chemistry, fluid inclusions and stable isotopes. Ore Geol Rev 117:103269\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHe H, Zhu H, Yang X, Zhang W, Wang J (2024) Mineral prospectivity prediction based on convolutional neural network and ensemble learning. Scientific Reports 2024 14:1 14, 1\u0026ndash;21. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s41598-024-73357-0\u003c/span\u003e\u003cspan address=\"10.1038/s41598-024-73357-0\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHezarkhani A (2006) Petrology of the intrusive rocks within the Sungun porphyry copper deposit, Azerbaijan, Iran. J Asian Earth Sci 27(3):326\u0026ndash;340\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHezarkhani A (2008) Hydrothermal evolution of the Miduk Porphyry copper system, Kerman, Iran: A fluid inclusion investigation. Int Geol Rev 50:665\u0026ndash;684. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.2747/0020-6814.50.7.665\u003c/span\u003e\u003cspan address=\"10.2747/0020-6814.50.7.665\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHolden EJ, Dentith M, Kovesi P (2008) Towards the automated analysis of regional aeromagnetic data to identify regions prospective for gold deposits. Comput Geosci 34:1505\u0026ndash;1513. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/J.CAGEO.2007.08.007\u003c/span\u003e\u003cspan address=\"10.1016/J.CAGEO.2007.08.007\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHronsky JMA, Groves DI (2008) Science of targeting: Definition, strategies, targeting and performance measurement. Aust J Earth Sci 55:3\u0026ndash;12. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/08120090701581356;WGROUP:STRING:PUBLICATION\u003c/span\u003e\u003cspan address=\"10.1080/08120090701581356;WGROUP:STRING:PUBLICATION\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eIrandoust MA, Priestley K, Sobouti F (2022) High-Resolution Lithospheric Structure of the Zagros Collision Zone and Iranian Plateau. J Geophys Res Solid Earth 127. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1029/2022JB025009\u003c/span\u003e\u003cspan address=\"10.1029/2022JB025009\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eJamali H, Mehrabi B (2015) Relationships between arc maturity and Cu-Mo-Au porphyry and related epithermal mineralization at the Cenozoic Arasbaran magmatic belt. Ore Geol Rev 65:487\u0026ndash;501. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/J.OREGEOREV.2014.06.017\u003c/span\u003e\u003cspan address=\"10.1016/J.OREGEOREV.2014.06.017\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKesler SE, Simon AC (2015) Mineral Resources, Economics and the Environment. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.7302/22482\u003c/span\u003e\u003cspan address=\"10.7302/22482\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKheyrollahi H, Alinia F, Ghods A (2018) Regional magnetic lithologies and structures as controls on porphyry copper deposits: Evidence from Iran. Explor Geophys 49:98\u0026ndash;110. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1071/EG16042\u003c/span\u003e\u003cspan address=\"10.1071/EG16042\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKnox-Robinson CM, Wyborn LAI (1997) Towards a holistic exploration strategy: Using Geographic Information Systems as a tool to enhance exploration. Aust J Earth Sci 44:453\u0026ndash;463. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/08120099708728326\u003c/span\u003e\u003cspan address=\"10.1080/08120099708728326\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKorehie MT, Ardebili O, Ahari HD, Shirzad MR, Fotovati V, Ghalamghash J, Kiani T, Najafi A, Soltani NS, Ashtiani ME, Farhatjah B (2019) Atlas of Iran\u0026rsquo;s Geology and Mineral Distribution. Springer, Springer Nature\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLeclerc F, Harris LB, B\u0026eacute;dard JH, Van Breemen O, Goulet N (2012) Structural and stratigraphic controls on magmatic, volcanogenic, and shear zone-hosted mineralization in the Chapais-Chibougamau mining camp, northeastern Abitibi, Canada (1,2). Econ Geol 107:963\u0026ndash;969. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.2113/ECONGEO.107.5.963\u003c/span\u003e\u003cspan address=\"10.2113/ECONGEO.107.5.963\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLi Q, Chen G, Wang D (2024) Mineral Prospectivity Mapping Using Semi-supervised Machine Learning. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/S11004-024-10161-6/FIGURES/16\u003c/span\u003e\u003cspan address=\"10.1007/S11004-024-10161-6/FIGURES/16\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. Math Geosci 1\u0026ndash;31\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLi S, Chen J, Liu C, Wang Y (2021) Mineral prospectivity prediction via convolutional neural networks based on geological big data. J Earth Sci 32:327\u0026ndash;347. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s12583-020-1365-z\u003c/span\u003e\u003cspan address=\"10.1007/s12583-020-1365-z\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLi YE, O\u0026rsquo;Malley D, Beroza G, Curtis A, Johnson P (2023) Machine Learning Developments and Applications in Solid-Earth Geosciences: Fad or Future? J Geophys Res Solid Earth 128. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1029/2022JB026310\u003c/span\u003e\u003cspan address=\"10.1029/2022JB026310\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eL\u0026ouml;sing M, Ebbing J (2021) Predicting Geothermal Heat Flow in Antarctica With a Machine Learning Approach. J Geophys Res Solid Earth 126:1\u0026ndash;16. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1029/2020JB021499\u003c/span\u003e\u003cspan address=\"10.1029/2020JB021499\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLou Y, Liu Y (2023) Mineral Prospectivity Mapping of Tungsten Polymetallic Deposits Using Machine Learning Algorithms and Comparison of Their Performance in the Gannan Region, China. Earth Space Sci 10. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1029/2022EA002596\u003c/span\u003e\u003cspan address=\"10.1029/2022EA002596\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. e2022EA002596\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMaepa F, Smith RS, Tessema A (2021) Support vector machine and artificial neural network modelling of orogenic gold prospectivity mapping in the Swayze greenstone belt, Ontario, Canada. Ore Geol Rev 130:103968. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.oregeorev.2020.103968\u003c/span\u003e\u003cspan address=\"10.1016/j.oregeorev.2020.103968\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMaghfouri S, Hosseinzadeh MR (2018) The early Cretaceous Mansourabad shale-carbonate hosted Zn-Pb (-Ag) deposit, central Iran: An example of vent-proximal sub-seafloor replacement SEDEX mineralization. Ore Geol Rev 95:20\u0026ndash;39\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMaghfouri S, Rastad E, Momenzadeh M, Movahednia M, Hashempour SS, Jadehkenary KA, Ghaderi M, Konari MB, Izanloo J, Yarmohammadi A, Peernajmodin H (2025) Malayer-Esfahan Metallogenic Belt, Iran: Jurassic-Early Cretaceous sediment-(volcanic) hosted Zn-Pb (\u0026plusmn;\u0026thinsp;Ba\u0026thinsp;\u0026plusmn;\u0026thinsp;Ag), Fe-Mn-Pb (\u0026plusmn;\u0026thinsp;Ba\u0026thinsp;\u0026plusmn;\u0026thinsp;Cu) and Ba (\u0026plusmn;\u0026thinsp;Pb\u0026thinsp;\u0026plusmn;\u0026thinsp;Zn) deposits with evolution of basin. Journal of the Geological Society, jgs2024-284.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMahboob MA, Celik T, Genc B (2024) Predictive modelling of mineral prospectivity using satellite remote sensing and machine learning algorithms. Remote Sens Appl 36:101316. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/J.RSASE.2024.101316\u003c/span\u003e\u003cspan address=\"10.1016/J.RSASE.2024.101316\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMateus A, Martins L (2019) Challenges and opportunities for a successful mining industry in the future. Bolet\u0026iacute;n Geol\u0026oacute;gico y Min 130:99\u0026ndash;121. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.21701/BOLGEOMIN.130.1.007\u003c/span\u003e\u003cspan address=\"10.21701/BOLGEOMIN.130.1.007\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMcCuaig TC, Beresford S, Hronsky J (2010) Translating the mineral systems approach into an effective exploration targeting system. Ore Geol Rev 38:128\u0026ndash;138. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/J.OREGEOREV.2010.05.008\u003c/span\u003e\u003cspan address=\"10.1016/J.OREGEOREV.2010.05.008\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMcCuaig TC, Hronsky J (2017) The mineral systems concept: the key to exploration targeting. Appl Earth Sci 126:77\u0026ndash;78. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/03717453.2017.1306274\u003c/span\u003e\u003cspan address=\"10.1080/03717453.2017.1306274\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMcMillan M, Haber E, Peters B, Fohring J (2021) Mineral prospectivity mapping using a VNet convolutional neural network. Lead Edge 40:99\u0026ndash;105. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1190/tle40020099.1\u003c/span\u003e\u003cspan address=\"10.1190/tle40020099.1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMehrdar A, Motaghi K, Ghods A, Sobouti F, Priestley K, Pachhai S, Shabanian E, Zarunizadeh Z, Zeynaddini-Meymand R, El-Hussain I (2025) Crustal and uppermost mantle structure of the Iranian Makran subduction zone from ambient noise and earthquake surface wave tomography. Geophys J Int 241:70\u0026ndash;85. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1093/GJI/GGAE419\u003c/span\u003e\u003cspan address=\"10.1093/GJI/GGAE419\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMeshkani SA, Mehrabi B, Yaghubpur A, Sadeghi M (2013) Recognition of the regional lineaments of Iran: Using geospatial data and their implications for exploration of metallic ore deposits. Ore Geol Rev 55:48\u0026ndash;63. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/J.OREGEOREV.2013.04.007\u003c/span\u003e\u003cspan address=\"10.1016/J.OREGEOREV.2013.04.007\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMehrabi B, Siani MG, Goldfarb R, Azizi H, Ganerod M, Marsh EE (2016) Mineral assemblages, fluid evolution, and genesis of polymetallic epithermal veins, Glojeh district, NW Iran. Ore Geol Rev 78:41\u0026ndash;57\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMollai H, Sharma R, Pe-Piper G (2009) Copper mineralization around the Ahar batholith, north of Ahar (NW Iran): Evidence for fluid evolution and the origin of the skarn ore deposit. Ore Geol Rev 35:401\u0026ndash;414. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.oregeorev.2009.02.005\u003c/span\u003e\u003cspan address=\"10.1016/j.oregeorev.2009.02.005\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMonsef I, Monsef R, Mata J, Zhang Z, Pirouz M, Rezaeian M, Esmaeili R, Xiao W (2018) Evidence for an early-MORB to fore-arc evolution within the Zagros suture zone: Constraints from zircon U-Pb geochronology and geochemistry of the Neyriz ophiolite (South Iran). Gondwana Res 62:287\u0026ndash;305. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/J.GR.2018.03.002\u003c/span\u003e\u003cspan address=\"10.1016/J.GR.2018.03.002\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMonsef I, Rahgoshay M, Pirouz M, Chiaradia M, Gr\u0026eacute;goire M, Ceuleneer G (2019) The Eastern Makran Ophiolite (SE Iran): evidence for a Late Cretaceous fore-arc oceanic crust. Int Geol Rev 61:1313\u0026ndash;1339. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/00206814.2018.1507764\u003c/span\u003e\u003cspan address=\"10.1080/00206814.2018.1507764\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMonsef I, Zhang Z, Shabanian E, le Roux P, Rahgoshay M (2022) Tethyan subduction and Cretaceous rift magmatism at the southern margin of Eurasia: Evidence for crustal evolution of the South Caspian Basin. Earth Sci Rev 228:104012. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/J.EARSCIREV.2022.104012\u003c/span\u003e\u003cspan address=\"10.1016/J.EARSCIREV.2022.104012\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMoritz R, Ghazban F, Singer BS (2006) Eocene gold ore formation at Muteh, Sanandaj-Sirjan tectonic zone, western Iran: A result of late-stage extension and exhumation of metamorphic basement rocks within the Zagros orogen. Econ Geol 101:1497\u0026ndash;1524. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.2113/GSECONGEO.101.8.1497\u003c/span\u003e\u003cspan address=\"10.2113/GSECONGEO.101.8.1497\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMousivand F, Rastad E, Peter JM, Maghfouri S (2018) Metallogeny of volcanogenic massive sulfide deposits of Iran. Ore Geol Rev 95:974\u0026ndash;1007\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eNabatian G, Rastad E, Neubauer F, Honarmand M, Ghaderi M (2015) Iron and Fe\u0026ndash;Mn mineralisation in Iran: implications for Tethyan metallogeny. Aust J Earth Sci 62:211\u0026ndash;241. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/08120099.2015.1002001\u003c/span\u003e\u003cspan address=\"10.1080/08120099.2015.1002001\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eOmidianfar S, Monsef I, Rahgoshay M, Zheng J, Cousens B (2020) The middle Eocene high-K magmatism in Eastern Iran Magmatic Belt: constraints from U-Pb zircon geochronology and Sr-Nd isotopic ratios. Int Geol Rev 62:1751\u0026ndash;1768. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/00206814.2020.1716272\u003c/span\u003e\u003cspan address=\"10.1080/00206814.2020.1716272\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eParsa M, Carranza EJM (2021) Modulating the impacts of stochastic uncertainties linked to deposit locations in data-driven predictive mapping of mineral prospectivity. Nat Resour Res 30:3081\u0026ndash;3097. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s11053-021-09891-9\u003c/span\u003e\u003cspan address=\"10.1007/s11053-021-09891-9\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eParsa M, Maghsoudi A (2021) Assessing the effects of mineral systems-derived exploration targeting criteria for random Forests-based predictive mapping of mineral prospectivity in Ahar-Arasbaran area, Iran. Ore Geol Rev 138:104399. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.oregeorev.2021.104399\u003c/span\u003e\u003cspan address=\"10.1016/j.oregeorev.2021.104399\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003ePavlis NK, Holmes SA, Kenyon SC, Factor JK (2012a) The development and evaluation of the Earth Gravitational Model 2008 (EGM2008). J Geophys Res Solid Earth 117:1\u0026ndash;38. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1029/2011JB008916\u003c/span\u003e\u003cspan address=\"10.1029/2011JB008916\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003ePavlis NK, Holmes SA, Kenyon SC, Factor JK (2012b) The development and evaluation of the Earth Gravitational Model 2008 (EGM2008). J Geophys Res Solid Earth 117:1\u0026ndash;38. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1029/2011JB008916\u003c/span\u003e\u003cspan address=\"10.1029/2011JB008916\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003ePenney C, Tavakoli F, Saadat A, Nankali HR, Sedighi M, Khorrami F, Sobouti F, Rafi Z, Copley A, Jackson J, Priestley K (2017) Megathrust and accretionary wedge properties and behaviour in the Makran subduction zone. Geophys J Int 209:1800\u0026ndash;1830. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1093/gji/ggx126\u003c/span\u003e\u003cspan address=\"10.1093/gji/ggx126\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003ePetrella L, Williams-Jones AE, Goutier J, Walsh J (2014) The nature and origin of the rare earth element mineralization in the misery syenitic intrusion, Northern Quebec, Canada. Econ Geol 109:1643\u0026ndash;1666. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.2113/ECONGEO.109.6.1643\u003c/span\u003e\u003cspan address=\"10.2113/ECONGEO.109.6.1643\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eQaderi S, Maghsoudi A, Pour AB, Yousefi M (2024) Geological Controlling Factors on Mississippi Valley-Type Pb-Zn Mineralization in Western Semnan, Iran. Minerals (2075-163X), 14(9).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eRajabi A, Rastad E, Canet C (2013) Metallogeny of Permian-Triassic carbonate-hosted Zn-Pb and F deposits of Iran: A review for future mineral exploration. Aust J Earth Sci 60(2):197\u0026ndash;216\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eRichards JP (2015) Tectonic, magmatic, and metallogenic evolution of the Tethyan orogen: From subduction to collision. Ore Geol Rev 70:323\u0026ndash;345. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/J.OREGEOREV.2014.11.009\u003c/span\u003e\u003cspan address=\"10.1016/J.OREGEOREV.2014.11.009\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eRichards JP, Wilkinson D, Ullrich T (2006) Geology of the Sari Gunay epithermal gold deposit, northwest Iran. Econ Geol 101:1455\u0026ndash;1496. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.2113/GSECONGEO.101.8.1455\u003c/span\u003e\u003cspan address=\"10.2113/GSECONGEO.101.8.1455\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eRodriguez-Galiano V, Sanchez-Castillo M, Chica-Olmo M, Chica-Rivas M, Machine learning predictive models for mineral prospectivity: An evaluation of neural networks, random forest, regression trees and support vector machines. Ore Geol Rev 71, 804\u0026ndash;818., Sahandi M, Soheili M (2015) 2014. Geological Map of Iran, scale 1:1000000. Geological Survey of Iran\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eShahabpour J (1999) The role of deep structures in the distribution of some major ore deposits in Iran, NE of the Zagros thrust zone. J Geodyn 28:237\u0026ndash;250. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/S0264-3707(98)00040-4\u003c/span\u003e\u003cspan address=\"10.1016/S0264-3707(98)00040-4\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eShayeganpour S, Tangestani MH (2022) Extraction of rock and alteration geons by FODPSO segmentation and GP regression on the HyMap imagery: A case study of SW Birjand, Eastern Iran. Ore Geol Rev 143:104767. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/J.OREGEOREV.2022.104767\u003c/span\u003e\u003cspan address=\"10.1016/J.OREGEOREV.2022.104767\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eShirmard H, Farahbakhsh E, M\u0026uuml;ller RD, Chandra R (2022) A review of machine learning in processing remote sensing data for mineral exploration. Remote Sens Environ 268. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.rse.2021.112750\u003c/span\u003e\u003cspan address=\"10.1016/j.rse.2021.112750\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSinger DA (2007) Estimating amounts of undiscovered mineral resources. In Proceedings for a Workshop on Deposit Modeling, Mineral Resource Assessment, and Their Role in Sustainable Development: USGS Circular. (Vol. 1294, pp. 79\u0026ndash;84)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSoloviev SG, Kryazhev SG, Dvurechenskaya SS, Vasyukov VE, Shumilin DA, Voskresensky KI (2019) The superlarge Malmyzh porphyry Cu-Au deposit, Sikhote-Alin, eastern Russia: Igneous geochemistry, hydrothermal alteration, mineralization, and fluid inclusion characteristics. Ore Geol Rev 113. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.oregeorev.2019.103112\u003c/span\u003e\u003cspan address=\"10.1016/j.oregeorev.2019.103112\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eStampfli GM (2000) Tethyan oceans. Geological Society, London, Special Publications 173, 1\u0026ndash;23. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1144/GSL.SP.2000.173.01.01\u003c/span\u003e\u003cspan address=\"10.1144/GSL.SP.2000.173.01.01\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eStampfli GM, Borel GD (2004) The TRANSMED Transects in Space and Time: Constraints on the Paleotectonic Evolution of the Mediterranean Domain. Springe, Berlin, pp 53\u0026ndash;90\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eStampfli GM, Borel GD (2002) A plate tectonic model for the Paleozoic and Mesozoic constrained by dynamic plate boundaries and restored synthetic oceanic isochrons. Earth Planet Sci Lett 196:17\u0026ndash;33. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/S0012-821X(01)00588-X\u003c/span\u003e\u003cspan address=\"10.1016/S0012-821X(01)00588-X\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSt\u0026ouml;cklin J (1974) Possible Ancient Continental Margins in Iran. The Geology of Continental Margins. Springer, Berlin Heidelberg, pp 873\u0026ndash;887. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/978-3-662-01141-6_64\u003c/span\u003e\u003cspan address=\"10.1007/978-3-662-01141-6_64\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eStocklin J (1968) Structural History and Tectonics of Iran: A Review. Am Assoc Pet Geol Bull 52:1229\u0026ndash;1258\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSt\u0026ouml;cklin J (1968) Structural history and tectonics of Iran: a review. Am Assoc Pet Geol Bull 52:1229\u0026ndash;1258\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eStosch HG, Romer RL, Daliran F, Rhede D (2011) Uranium-lead ages of apatite from iron oxide ores of the Bafq District, East-Central Iran. Min Depos 46:9\u0026ndash;21. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/S00126-010-0309-4\u003c/span\u003e\u003cspan address=\"10.1007/S00126-010-0309-4\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSun T, Li H, Wu K, Chen F, Zhu Z, Hu Z (2020) Data-driven predictive modelling of mineral prospectivity using machine learning and deep learning methods: a case study from southern Jiangxi Province China. Minerals 10:102. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/min10020102\u003c/span\u003e\u003cspan address=\"10.3390/min10020102\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eTagwai MG, Jimoh OA, Shehu SA, Zabidi H (2024) Application of GIS and remote sensing in mineral exploration: current and future perspectives. World J Eng 21:487\u0026ndash;502. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1108/WJE-09-2022-0395\u003c/span\u003e\u003cspan address=\"10.1108/WJE-09-2022-0395\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eTaleFazel E, Mehrabi B, GhasemiSiani M (2019) Epithermal systems of the Torud-Chah Shirin district, northern Iran: Ore-fluid evolution and geodynamic setting. Ore Geol Rev 109:253\u0026ndash;275\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eTeknik V (2024) The improved Moho depth imaging in the Arabia-Eurasia collision zone: A machine learning approach integrating seismic observations and satellite gravity data. Tectonophysics 893:230553. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.tecto.2024.230553\u003c/span\u003e\u003cspan address=\"10.1016/j.tecto.2024.230553\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eTeknik V, Artemieva IM, Thybo H (2024) Limited arc magmatism and seismicity due to extensive mantle wedge serpentinization in the Makran subduction zone. Earth Planet Sci Lett 645:118950. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.epsl.2024.118950\u003c/span\u003e\u003cspan address=\"10.1016/j.epsl.2024.118950\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eTeknik V, Ghods A (2017) Depth of magnetic basement in Iran based on fractal spectral analysis of aeromagnetic data. Geophys J Int 209:1878\u0026ndash;1891. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1093/gji/ggx132\u003c/span\u003e\u003cspan address=\"10.1093/gji/ggx132\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eTeknik V, Thybo H, Artemieva IM, Ghods A (2020) A new tectonic map of the Iranian plateau based on aeromagnetic identification of magmatic arcs and ophiolite belts. Tectonophysics 792. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/J.TECTO.2020.228588\u003c/span\u003e\u003cspan address=\"10.1016/J.TECTO.2020.228588\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eTorab FM, Lehmann B (2007) Magnetite-apatite deposits of the Bafq district, Central Iran: apatite geochemistry and monazite geochronology. Mineral Mag 71:347\u0026ndash;363. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1180/MINMAG.2007.071.3.347\u003c/span\u003e\u003cspan address=\"10.1180/MINMAG.2007.071.3.347\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eVerdel C, Wernicke BP, Hassanzadeh J, Guest B (2011) A Paleogene extensional arc flare - up in Iran. Tectonics 30:1\u0026ndash;20. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1029/2010TC002809\u003c/span\u003e\u003cspan address=\"10.1029/2010TC002809\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eWake N, Farahbakhsh E, M\u0026uuml;ller RD (2024) Lateritic Ni\u0026ndash;Co Prospectivity Modeling in Eastern Australia Using an Enhanced Generative Adversarial Network and Positive-Unlabeled Bagging. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/S11053-024-10423-4\u003c/span\u003e\u003cspan address=\"10.1007/S11053-024-10423-4\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. Natural Resources Research\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eWoodhead J, Landry M (2021) Harnessing the Power of Artificial Intelligence and Machine Learning in Mineral Exploration\u0026mdash;Opportunities and Cautionary Notes. SEG Discovery 19\u0026ndash;31. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.5382/Geo-and-Mining-13\u003c/span\u003e\u003cspan address=\"10.5382/Geo-and-Mining-13\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eXiong Y, Zuo R (2020) Recognizing multivariate geochemical anomalies for mineral exploration by combining deep learning and one-class support vector machine. Comput Geosci 140. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.cageo.2020.104484\u003c/span\u003e\u003cspan address=\"10.1016/j.cageo.2020.104484\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eXiong Y, Zuo R (2018) GIS-based rare events logistic regression for mineral prospectivity mapping. Comput Geosci 111:18\u0026ndash;25. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.cageo.2017.10.005\u003c/span\u003e\u003cspan address=\"10.1016/j.cageo.2017.10.005\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eXiong Y, Zuo R, Carranza EJM (2018) Mapping mineral prospectivity through big data analytics and a deep learning algorithm. Ore Geol Rev 102:811\u0026ndash;817. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.oregeorev.2018.10.006\u003c/span\u003e\u003cspan address=\"10.1016/j.oregeorev.2018.10.006\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eYang H, huan, Wang Q, Li Ybo, Lin B, Song Y, Wang Y, yun, He W, Li H, wei, Li S, Li J, li, Liu C, cheng, Feng S, bin, Xin T, Fu Xlian, Liang X, juan, Zhang Q, Wang Bqi, Li Y (2022) Geology and mineralization of the Tiegelongnan supergiant porphyry-epithermal Cu (Au, Ag) deposit (10 Mt) in western Tibet, China: A review. China Geol 5:136\u0026ndash;159. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/S2096-5192(22)00091-X\u003c/span\u003e\u003cspan address=\"10.1016/S2096-5192(22)00091-X\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eYang N, Zhang Z, Yang J, Hong Z (2022) Mineral prospectivity prediction by integration of convolutional autoencoder network and random forest. Nat Resour Res 31:1103\u0026ndash;1119. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s11053-022-10038-7\u003c/span\u003e\u003cspan address=\"10.1007/s11053-022-10038-7\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eYin B, Zuo R, Sun S (2023) Mineral prospectivity mapping using deep self-attention model. Nat Resour Res 32:37\u0026ndash;56. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s11053-022-10142-8\u003c/span\u003e\u003cspan address=\"10.1007/s11053-022-10142-8\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eYin J, Li N (2022) Ensemble learning models with a Bayesian optimization algorithm for mineral prospectivity mapping. Ore Geol Rev 145. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.oregeorev.2022.104916\u003c/span\u003e\u003cspan address=\"10.1016/j.oregeorev.2022.104916\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZarasvandi A, Liaghat S, Zentilli M (2005) Geology of the Darreh-Zerreshk and Ali-Abad porphyry copper deposits, central Iran. Int Geol Rev 47(6):620\u0026ndash;646\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZarasvandi A, Liaghat S, Zentilli M, Reynolds PH (2007) 40Ar/39Ar geochronology of alteration and petrogenesis of porphyry copper-related granitoids in the Darreh-Zerreshk and Ali-Abad area, central Iran. Explor Min Geol 16:11\u0026ndash;24. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.2113/GSEMG.16.1-2.11\u003c/span\u003e\u003cspan address=\"10.2113/GSEMG.16.1-2.11\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZarasvandi A, Rezaei M, Sadeghi M, Lentz D, Adelpour M, Pourkaseb H (2015) Rare earth element signatures of economic and sub-economic porphyry copper systems in Urumieh-Dokhtar Magmatic Arc (UDMA), Iran. Ore Geol Rev 70:407\u0026ndash;423. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/J.OREGEOREV.2015.01.010\u003c/span\u003e\u003cspan address=\"10.1016/J.OREGEOREV.2015.01.010\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZarasvandi A, Rezaei M, Raith JG, Asadi S, Lentz D (2019) Hydrothermal fluid evolution in collisional Miocene porphyry copper deposits in Iran: Insights into factors controlling metal fertility. Ore Geol Rev 105:183\u0026ndash;200\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZhang SE, Lawley CJM, Bourdeau JE, Nwaila GT, Ghorbani Y (2024) Nat Resour Res 2024 33(3 33):995\u0026ndash;1023. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/S11053-024-10322-8\u003c/span\u003e\u003cspan address=\"10.1007/S11053-024-10322-8\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. Workflow-Induced Uncertainty in Data-Driven Mineral Prospectivity Mapping\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZhao P (2007) Quantitative mineral prediction and deep mineral exploration. Earth Sci Front 14:1\u0026ndash;10\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZuo R, Xu Y (2023) Graph deep learning model for mapping mineral prospectivity. Math Geosci 55:1\u0026ndash;21. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s11004-022-10015-z\u003c/span\u003e\u003cspan address=\"10.1007/s11004-022-10015-z\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZuo R, Peng Y, Li T, Xiong Y (2021) Challenges of geological prospecting big data mining and integration using deep learning algorithms. Earth Sci 46:350\u0026ndash;358\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Geospatial Big Data, Machine Learning, Mineral Prospectivity Mapping, Artificial Intelligence, Green exploration, Ore Deposits, Iran","lastPublishedDoi":"10.21203/rs.3.rs-7730637/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7730637/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eRecent advances in Artificial Intelligence (AI) and Machine Learning (ML) methods have significantly enhanced Mineral Prospectivity Mapping (MPM). These AI-based algorithms offer high capability for regional-scale mapping of underexplored ore deposits. However, there are still significant methodological challenges, particularly in integrating multidimensional, heterogeneous geospatial datasets and handling their inconsistencies with sparse and spatially clustered distributed known mineral ore deposits. The current study presents a novel framework to address these challenges. We developed a training dataset comprising 69 training features derived from geological, geophysical, and lithospheric raster grids. Vector-based geological features were systematically converted into raster grids, where each pixel encodes the minimum distance to the nearest structural and lithological boundaries. Therefore, one can capture the influence of structural and lithological proximity on metallogenic zones. The training target is generated by combining the spatial distribution of seven major metallic ore deposits and converting their spatial locations into a continuous raster grid of spatial density of metallic ore occurrence. Seven ML algorithms with their 24 subtypes are used to predict the spatial density of mineral deposits. Among them, the Ensemble Bagged Trees method showed optimum prediction performance by achieving the lowest Root Mean Square Error (RMSE) and the highest coefficient of determination (R\u0026sup2;). The optimized model was applied to calculate a predictive MPM across the Iranian plateau. To pinpoint underexplored high-potential zones, residual spatial density anomalies were calculated by subtracting the observed ore occurrence spatial densities from the predicted prospective grid. The residual spatial density anomalies reveal several promising areas, such as the Malayer-Isfahan Pb-Zn zone along the Zagros suture zone. The residual anomalies show significant potential extended southward of the KaraDagh copper zone in NW Iran. The results also indicate high potential zones in central and eastern Iran, notably near the Bafgh, Nehbandan-Ferdous, and Jiroft-Shahrebabak metallogenic zones. Regional-scale AI-aided regression analysis enhances our understanding of ore deposit distribution across the Iranian plateau. This insight provides a strategic foundation for future national-scale exploration programs by improving efficiency, reducing risk and cost, and narrowing the area of detailed exploration.\u003c/p\u003e","manuscriptTitle":"Machine learning (ML)-based mineral prospectivity mapping (MPM): Detecting Iranian plateau high-potential metallogenic zones using geospatial big data","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-09-30 08:17:43","doi":"10.21203/rs.3.rs-7730637/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"ef0f5b98-1d03-4861-a543-3c2d5115ff42","owner":[],"postedDate":"September 30th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":55442956,"name":"Artificial Intelligence and Machine Learning"},{"id":55442957,"name":"Geophysics"}],"tags":[],"updatedAt":"2025-09-30T08:17:43+00:00","versionOfRecord":[],"versionCreatedAt":"2025-09-30 08:17:43","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-7730637","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7730637","identity":"rs-7730637","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}